Build a Custom AI-Powered GPT3 Chatbot For Your Business
The Rise Of ChatGPT ChatGPT has rapidly gained popularity worldwide, with millions of users relying on its vast knowledge database and ability to hold a multi-turn conversation of considerable complexity. However, despite its usefulness for general information, ChatGPT is limited to pre-2021 publicly available internet data, and it has no access to your private data…Read More
What’s The Difference Between ChatGPT & GPT3?
ChatGPT Confusion At the time of writing, since the launch of ChatGPT at the end of November 2022, numerous solutions have hit the market claiming to be ChatGPT-branded. However, I’m here to clarify that these solutions are not ChatGPT, but rather GPT3 solutions. There seems to be a lot of confusion between ChatGPT and GPT3.…Read More
What is ChatGPT?
AI-Mazing ChatGPT is the latest technology release from the team at OpenAI and it’s taken the internet by storm. Reactions ranged from amazement to scepticism, with everything in between. It’s been hugely popular with more than a million people using it in the first 5 days. ChatGPT is OpenAI’s latest large language model, released on…Read More
Arabic NLP Guide [2023 Update]
Introduction Arabic is the fourth most spoken language on the internet and arguably one of the most difficult languages to create automated conversational experiences for, such as chatbots. An Arabic chatbot is a program that can understand and respond in Arabic. Natural language technologies enabling us to simulate and process human conversations in Arabic have improved…Read More
9 Questions To Help Define Your Chatbot Project Scope [2023 Update]
Defining Your Chatbot Project Scope So what is a scope of work? In terms of conversational AI development, a scope of work typically outlines the specific tasks and objectives that will be accomplished during the project. This can include details such as the functionality of the chatbot, the technologies and platforms to be used, the…Read More
How Much Does it Cost To Build a Chatbot in 2023?
Tips. Insight. Offers. Are You In? Your Email Address Please enter a valid email address. I agree that The Bot Forge can email me news, tips, updates & offers. I know that I can unsubscribe at any time. You must accept the Terms and Conditions. Sign Me Up Thank you for subscribing! Something went wrong.…Read More
The Bot Forge is One of the Most Reviewed UK AI Companies in 2022
The Manifest Recognizes The Bot Forge as One of the Most Reviewed AI Companies in the UK The Bot Forge creates modern solutions to improve organizational efficiency for our partners. Our team of experts helps you design, build, launch, or support enterprise-grade chatbots, voice assistants, and conversational IVR solutions. We aid you in making the…Read More
How To Create The Perfect Google Business Welcome Message
Why Google Business Messages? Google Maps has 155 million monthly users and it’s estimated that Google handles 5.6 billion searches per day – two trillion searches a year! Connecting with your customers at these two touchpoints is more beneficial than ever. Gone are the days when you needed to send customers to a website or social media profile…Read More
3 Books That Will Boost Your Chatbot Knowledge
Introduction Currently, chatbots are dominating online markets, especially in countries such as the U.S., India, Germany, Brazil, and the UK. According to a Business Insider article on chatbot statistics, 40% of internet users worldwide prefer chatbots over virtual agents because they get answers quickly and more conveniently due to their 24-hour service. Due to the…Read More
6 Tips to Ensure Your Chatbot is GDPR Compliant
Tips. Insight. Offers. Are You In? Your Email Address Please enter a valid email address. I agree that The Bot Forge can email me news, tips, updates & offers. I know that I can unsubscribe at any time. You must accept the Terms and Conditions. Sign Me Up Thank you for subscribing! Something went wrong.…Read More
What Makes a Successful Chatbot Project?
Tips. Insight. Offers. Are You In? Your Email Address Please enter a valid email address. I agree that The Bot Forge can email me news, tips, updates & offers. I know that I can unsubscribe at any time. You must accept the Terms and Conditions. Sign Me Up Thank you for subscribing! Something went wrong.…Read More
10 Questions To Ask When Planning a Chatbot Project
Tips. Insight. Offers. Are You In? Your Email Address Please enter a valid email address. I agree that The Bot Forge can email me news, tips, updates & offers. I know that I can unsubscribe at any time. You must accept the Terms and Conditions. Sign Me Up Thank you for subscribing! Something went wrong.…Read More
Building a Chatbot Using Amazon Lex
What is Amazon Lex Amazon Lex is a service by AWS for building conversational interfaces into any application using voice and text. Lex has quickly become popular among chatbots enthusiasts looking to leverage the technology which powers Alexa. Users can be up and running within minutes with no upfront costs. Amazon Lex has been in…Read More
6 Common Mistakes to Avoid When Developing a Chatbot
We are experts in developing chatbots so we know If you are looking to streamline certain operations of your business, developing a chatbot is a great way to go about it. After all, you can use technologies such as artificial intelligence (AI) and natural language processing (NLP) so that it can follow different types of…Read More
Buckinghamshire Business Festival Sponsor 2021
We are proud to be a Buckinghamshire Business Festival sponsor this year We are proud to be sponsoring the 2021 Buckinghamshire Business Festival, running from April 19th – 30th. The festival has been organised by Buckinghamshire Business First, with a packed schedule of events and opportunities to make new connections across the two weeks. Look…Read More
IBM Watson for Building Chatbots
IBM Watson
Developer tools that make it easy to incorporate conversation, language, and search into your applications. Watson gives you access to detailed developer resources that help you get started fast, including documentation and SDKs on GitHub.
There are several IBM Watson APIs available on the IBM Cloud. One of them is IBM Watson Assistant. Watson Assistant enables you to build apps that include natural language processing and structured conversation. The service provides an API which you can call from an app or website to hook into your chatbot.
Watson Assistant API can:
- Extract meaning from natural language
- Discover patterns in data sets
- Understand the "tone" of language
- Translate languages
- Convert text to speech and speech to text
- Perform text classification
- Build a virtual agent (chatbot)
Watson is more of an assistant. It knows when to seek the answer from the knowledge base, when to ask for clarity and when to lead yourself to the human. Watson Assistant can work in any cloud-allowing businesses to bring AI to their data and apps wherever they are.
IBM Watson Assistant is marketed as a solution for companies of any size who want to build their voice or touch-enabled virtual assistant.
To create chatbot using IBM Watson API is mandatory to have a IBM/Bluemix account to start and its free (Lite Version.) Chatbot is built using intents, entities and Watson Developer Cloud to interact with the chatbot.
When we compare IBM Watson with Dialogflow, there is a question, what is better?
If you need a competent Artificial Intelligence Software product for your company you must make time to examine a wide range of alternatives. Aside from the robust features, the software which is simple and intuitive is always the better product.
In 2019, according to some market research, the user satisfaction level for IBM Watson is at 99% while for Dialogflow is at 96%. Both bot frameworks have their pros and cons. Dialogflow and Watson Assistant provide a UI tool to design conversation flow logic for complex dialogues.
Dialogflow provides maybe an easier and quicker way to create a custom conversational AI bot, while IBM Watson offering are targeting more corporations and enterprise organizations. For those who start to learn how to build a chatbot, maybe is better to choose and begin with Dialogflow.
Watson conversation is expensive compared to Dialogflow, while development interface in Dialogflow could have been better. Dialogflow bot for website integration does not support buttons and links while Watson Assistant for web integrations supports buttons and links usage.
Watson Assistant and Dialogflow integrate with variety of other popular platforms and systems.
Watson is not a single thing. Watson is a collection of APIs that can be used to solve various challenges and Watson Assistant is part of it. Many senior developers think that today there's nothing on market like Watson Assistant.
With the proper expectations and in the proper hands, Watson's APIs can be used to do some really phenomenal stuff.
More about Watson Assistant you can read at official IBM website: https://www.ibm.com/cloud/watson-assistant/
Enterprise AI Chatbot Integrations
The chatbots we create at The Bot Forge can do anything. We talk a lot about the chatbots themselves, NLP, Entities, Sentiment Analysis, Machine Learning, Training; all the good stuff which we leverage to make the optimum chat experience for our clients
However, sometimes we don't cover what goes on under the hood to ensure your chatbot does exactly what you need it to do.
Our enterprise AI Chatbot solutions' flexible nature gives you the freedom to build and expand on it however you see fit. No matter how niche your use case is, the solution will make it possible.
So what do we mean by service integrations? In this case we mean what systems do you want your chatbot to talk to or interact with to get their job done. To do it's job an enterprise AI Chatbot may need to integrate with multiple existing systems: CRM, internal knowledge base or meeting booking system. It all depends on your use-case.
Enterprise Content Management
Help your suppliers, customers, vendors and internal stakeholders in finding relevant documents quickly. Our chatbots will integrate with any internal company database or third-party database for your end-user to have appropriate human-like responses. At the same time these types of integrations allow easy management of bot responses as you have control over the single source of truth.
CRM Applications
Deliver a better experience to your consumers by integrating your customer support chatbots with Hubspot, Salesforce CRM or Zendesk.
ERP Systems
Chatbots can extend the capabilities of your ERP systems and can change the way you have been doing business.
Appointment Systems
Chatbots integrate with most popular appointment software solutions to book meeting and schedule appointments.
Check out some of the enterprise AI chatbot integrations we are using today and the potential they unlock. There are hundreds of examples, stretching across all sectors, these are just a few. Get in touch if you have your own integration in mind.
Connect chatbots to Google Sheets so your chatbot can respond accurately by connecting to up to date information.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Google Launches New Dialogflow CX in Beta
Google has beta launched a new version of it's Dialogflow natural language understanding (NLU) platform: Dialogflow CX. The Dialogflow version we all know and love is now called Dialogflow ES.
The new Dialogflow Customer Experience (CX) platform is aimed at building advanced artificial intelligence agents for enterprise-level projects at a larger and more complex scale than the standard variety.
"Dialogflow CX provides a new way of designing agents, taking a state machine approach to agent design," Google explained in the documentation for CX." This gives you clear and explicit control over a conversation, a better end-user experience, and a better development workflow."
Stand Out Features
The stand out feature for us is the that the Interactive flow visualizations allow conversation builders to quickly see, understand, and edit their work so creating more complex multi turn transactional conversational experiences will be more straightforward.
A state-based data model allows developers to reuse intents, intuitively define transitions, and handle supplemental questions.
In a single virtual agent, separate flows let multiple teams work simultaneously.
Plus, there seems to be versioning and environment at the flow level with other features such as the ability to run AB experiments and split traffic. We've started to look at the new Dialogflow CX console and things look really interesting.
A full break-down is beyond the scope of this post; we will be getting back to you with a more detailed feature analysis in future.
You can read more about Dialogflow CX here
Introductory Video can be see here
Beta Limitations
Its worth keep in mind Dialogflow CX is in beta, so some important features are not implemented yet. The following features found in Dialogflow ES are not implemented for Dialogflow CX yet:
Any language other than English (en)
Integrations
Knowledge connectors
System entity extension
History
Training data import
First Impressions
Our first impressions are that this will be a major tool for creating complex conversational enquiry heavy chatbots without having to juggle context. So particularly IVR chatbots or text chatbots which need to serve more complex roles.
It's also important to mention that this is a beta release, so some important features are not implemented yet. The following features found in Dialogflow ES are not implemented for Dialogflow CX:
As Google technology partners we are really excited about this new version of Dialogflow; if you want to learn more about Dialogflow CX and how an advanced chatbot can help your company please contact us to discuss further.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Dialogflow Mega Agent
A Mega Agent..so what?!
Ok so you could argue that I need to get out more...but I was excited to notice yesterday that there is a new feature which has sneaked into the Dialogflow console. This is the concept of a Mega Agent. It's the ability to set an agent type to mega agent so that you can combine multiple agents into one single agent.
So why is this so important? At The Bot Forge, some of our Dialogflow agents can have 1000's of intents, particularly if they are providing an information service for a knowledge base. Unfortunately, the knowledge base functionality can be limiting as looked at in my post: Dialogflow Knowledge Connectors so it's often necessary to create one intent per FAQ to get the required accuracy and control. This can quickly use up an agents 2000 intent limit.
We have recently had to look at creating our own version of a mega agent. This was to be used in a website chatbot implementation which would serve as a gatekeeper to initial enquiries so that we could hand over a conversation to a specific chatbot overseeing a specific knowledge domain. So not really ideal and involving more middleware complexity particularly as we were planning to handle some sort of context between all the agents.
There are some caveats, its still one GCP project and there is a maximum of 10 sub-agents per mega agent.
A Quick look at Mega Agents
It’s also important to remember this feature is in beta! You can read more about setting up the new Mega Agent here. At the time of writing the link on the add agent page is incorrect.
I took a really quick look at the new mega agent functionality.
Adding a mega agent
Adding a Mega Agent is pretty straightforward, when you add a new agent then you just select the switch:
Your mega agents are then listed in the agent list:
Adding a sub-agent
Once the agent is selected then a Sub Agent button is enabled:
After selecting the sub-agents button I had already created a test agent to use as my sub-agent so I connected it.
When choosing adding sub-agents you can select an environment or whether to include or exclude the knowledge Base. There is also a handy link to the sub-agent:
My test agent was a simple default agent with one added intent:
Does_mega_agent_work with one training phrase "does mega agent work"
Testing it out
So far so good. Just to recap I have created a mega agent and another agent to act as my sub. So now for a test drive of my Mega Agent in the Dialogflow simulator
Unfortunately, I didn't get the result I hoped for:
This was obviously an IAM permissions issue so I figured probably something which I had not done. I went back to the information page and re-read the section: Set up roles
Basically, to interact with a mega agent in the Dialogflow simulator, the service account that is linked to your mega agent in the Dialogflow Console needs a role with detect intent access for all sub-agents. To achieve this I went to the IAM permissions page for the sub-agent and added the mega agent's service account email address as a member of the project with a role of Dialogflow API Client.
Going back to the simulator and trying out does mega agent again resulted in the correct response from the sub-agent!
Where to go from here with mega agents.
For me, this is a major step for chatbots which have big numbers of intents > 2000. Or where different teams need to manage a particular knowledge area for one chatbot subject, use-case or topic area.
This post has really only taken a quick view of the new Dialogflow mega agent functionality. In a later post, I want to investigate leveraging contexts between agents and use a more complex example.
There are still some areas which need work though. The biggest one which springs to mind is that the training pages area of the console for a mega agent needs to be able to support the concept of sub-agents to assign sub intents. It's still just a beta feature so hopefully, more to come!
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Create a customer support chatbot for your website.
Is your business a victim of poor customer support, slow response time, and high customer demands?
You can use a website chatbot to increase customer satisfaction and retention.
Many companies suffer from the same customer support demands:
At The Bot Forge we create chatbots to be used at a variety of different customer touchpoints: Website, Facebook Messenger, WhatsApp, Slack, Microsoft teams, Alexa or Google Home.
Your customers demand faster support.
As a rule, people don’t like waiting. Keeping people waiting to get initial help can be very damaging to customer experience and directly influence customer retention.
Speed of response and speed of resolution are seen as the most important aspects of the
customer service experience, regardless of channel.
Long queue of customers waiting?
In order to help a customer effectively, your support agent can speak to at most 2 users at a time.
Many queries still come in via email which can be time consuming and often sit unanswered over long periods of time.
These types of support queries can often be handled by a website chatbot which can tap into your company knowledge base and provide support 24/7
You are spending 1000s of pounds in customer support?
You have a skilled but overstretched customer support team? Or you have outsourced it to an agency which doesn’t even understand your business or products?
You are struggleing to deal with the more complex queries because your staff are bogged down with simple customer questions which could be handled by a website chatbot.
Generate some leads whilst you provide effective customer support.
Are you looking for new ways to generate leads, turn website visitors into customers without annoying them?
In an oversaturated market, it’s best to bank on customer service as a predictor of
customer loyalty. Companies that invest in a good support not only gain through
increased loyalty and more successful upsells, but also through new customers who
are willing to pay more for a better onboarding experience.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
What Can We Expect From Conversational AI in 2021?
The use of conversational AI will continue to rise
Yes we are going to see continued growth in conversational AI in 2021. It's predicted that 1.4 billion people will use chatbots on a regular basis with $5 billion projected to be invested in chatbots by 2021. Voice assistant use will also grow. The use of voice assistants is expected to triple by 2023 (juniper research) and 50% of businesses will spend more on conversational than mobile in 2021.
Conversational commerce
For retail and e-commerce firms, conversational commerce, which is e-commerce transactions made by conversational methods such as texting and messaging, is generating waves. C-commerce not only allows brands to better serve their client base, but it’s opening up doors to new customers as well. A report made by Facebook states that 40% of global respondents said that c-commerce was their first introduction to online shopping.
97% of all respondents said that they plan to continue or increase their c-commerce spending in the future. Brands are likely to start considering how to leverage this trend and integrate messaging apps within their sales and marketing strategies.
Voice Commerce
According to Techopedia, Voice Commerce describes the utilization of voice recognition technology that enables consumers to purchase online merchandise or services.
Basically, it lets consumers buy products or services by simply using their voice. However Voice Commerce can also be part of a much wider customer journey, the transaction may not have to occur via voice. For example a consumer might have seen an ad for a product and asks Alexa about its price. The user then decides to buy it a few days later on the Amazon website. That’s why Voice Commerce involves much more than an isolated transaction process via voice.
Voice is a big deal, the number of digital voice assistants in use worldwide is estimated to reach 8 billion by 2023.
Already smart speaker users:
- research products
- add items to their shopping list
- track a package
- make a purchase
- provide ratings or reviews
- contact support
- reorder items
In 2021 there will be more Alexa in skill or Google Home action purchases as more retailers will leverage this medium; British supermarket chain Ocado has led this by example.
There will also be a continued rise in the enablement of product purchases:
Amazon has put a lot of time and effort into creating a seamless customer experience with its Echo devices.
Conversational AI taking the next steps
2021 will see a transformation for conversational AI chatbot capabilities with projects such as https://www.kuki.ai , Blenderbot , Meena and GPT-3. Open-domain chatbots will push the boundaries of what is possible. There will be an increase in AI chatbots that are personalised, processes more advanced problems and has a greater understanding of customer sentiment. In this way, your standard chatbots are likely to be replaced by conversational AI chatbots that are able to have a more human-like back and forth conversation.
These new technologies include very large language models: The Meena model has 2.6 billion parameters and is trained on 341 GB of text (1) so these models make huge computational demands. In 2021 as compute power continues to drop in price there will be a rise in availability of this open-domain chatbot technology.
Companion systems
As we live under the constraints inflicted by a global pandemic, we have been tackling an unexpected increase in alienation and loneliness in 2020. The demand to fulfil a companion role for AI assistants is something we expect to expand in 2021. With AI advancements this is becoming more realistic.
In 2020 chatbots took an informational role in many areas of the crisis; we covered a Covid support chatbot back in April. This looks set to continue in 2021.
Chatbots in immersive game experiences
Conversational AI technology looks set to be used in some really interesting ways in 2021. Particularly embedded in real-time games and integrated in multiple platforms.
Voice interaction will augment user interfaces.
We see a rise in the popularity of adding voice capabilities to software products. Specifically leveraging this sort of technology in touch screen situations. Software developers will improve their products by removing friction from the touch screen experience by bringing in voice controls. This sort of feature would be particularly useful for more complex search screens.
Conversational search
Voice search is now a rapidly growing form of access to information, but to be even more useful, it will need to become more conversational. Multiple conversational turns, follow-up on search responses, clarification and refining searches – are all aspects of natural conversation that Conversational AI is starting to replicate. These will be assisted by advancements in features such as Continued Conversation. At the same time voice search data and your own "voice" presence will become more important. Hey Google, who are "insert company name here".
Chatbots as sales assistants
In 2020 we have seen a rise in chatbots taking on the role of sales assistants. Providing specific knowledge about products is where this type of technology can excel: Providing product recommendations based on provided parameters. We've been working on these types of projects ourselves and will have more to show in 2021!
Smart IVR use will continue to grow
Speech recognition and natural language understanding for automated inbound and outbound request processing will rise. With companies offering advanced audio gateways and services such as Audiocodes. More and more legacy IVRs will be replaced with conversational IVRs: no more struggling with keypad input and overly complicated menu prompts. Advanced features such as automatic handover to live agents and multi voice options to give your Smart IVR a voice that matches your brand will improve the customer experience.
We will also see chatbot technology being utilised in different ways. Particularly in an Agent assist role, where chatbots will listen to call center conversations and provide advice, information or even responses to operatives in real-time.
Conclusion
2021 looks set to be an exciting year. Advancements in technology and changes in customer patterns and the workplace in many industries will continue to drive the growth and use of conversational AI. Everyone at The Bot Forge is looking forward to some really exciting projects in the new year!
(1) https://ai.googleblog.com/2020/01/towards-conversational-agent-that-can.html)
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
The Bot Forge Named UK Industry Leader
The Bot Forge named the leading chatbot and voice assistant agency in the UK by Clutch
Artificial Intelligence is a staple plot device in the sci-fi genre, often featured in a negative light. But in reality, predictive software improves the quality of life and leverages businesses to be more efficient and agile. No doubt AI technology can have a high-risk and high reward situation. If done correctly, its potential is unlimited. On the spot programming, automated customer support, and predictive analytics are just a few of its remarkable features.
We at The Bot Forge understand the capabilities software development and AI technology can bring to your organization. With bespoke chatbot and voice assistants as our core service, we can build the voice of your company to interact with your customers with no worries. Our process is guaranteed to make your lives, as well as your clients’ easier.
It is with great honor to announce that The Bot Forge has been chosen as one of the top AI companies in the UK. Our company is among the best on Clutch and it’s all thanks to the support of our esteemed clients.
"We are really excited to have been chosen as one of the leading chatbot and voice assistant agencies in the UK by Clutch." - Adrian Thompson, Founder of The Bot Forge
We are grateful to be recognized as an industry leader. This award and our 5-star rating wouldn’t have been possible with our clients!
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
The Bot Forge Named a Leader in AI
Did you know that more than 100,000 businesses are using chatbots to help optimize their customer experience?
Customers want instant replies, and chatbots are the way to achieve this, according to a 2018 Forbes article.
Here at The Bot Forge, we have been providing custom software development and AI services since 2018.
What they say
After working with many clients in many industries, we are thrilled to announce that Clutch, a B2B ratings and reviews firm, has listed us as one of the leading AI companies in the UK.
Additionally, we are on Clutch’s Leaders Matrix for top AI developers in the UK. The Leaders Matrix shows companies that are at the top of their targeted markets. The Bot Forge is one of the nine leaders on the Matrix.
We could not have received this recognition without our clients. We have worked with small and mid-market businesses, and these businesses represent a variety of industries.
The industries they are in include the business services, financial services, and IT industries.
We received a 5-star rating from Stitch AI, a digital engagement solutions company. We provided web development services to the company; initially, Stitch AI needed assistance in building a web portal where it could create advanced lead generation chatbots for any industry vertical.
We created a platform that helps the client manage its customers’ chatbots, and we continuously work with the client. The client has been happy with the quality of our work.
“...we’re happy with their work, and they’ve fixed any bugs in a timely manner."
— Managing Director, Stitch AI
Our Vision
At The Bot Forge, we are committed to our clients’ satisfaction. Our clients make us who we are
"Our vision is for our agency to become a global champion in creating custom chatbot solutions for our customers," said Adrian Thompson, founder of The Bot Forge.
Clutch’s sister site, The Manifest, which serves as a guide for businesses, also listed us as one of the top AI developers in the UK.
You can also see us on Visual Objects, Clutch’s portfolio-sharing sister site that features us on its list of top software developers.
Let us help your company revamp its customer experience.
Visit our Clutch profile.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Dialogflow CX Now Has a Free Trial
It was pretty exciting when Dialogflow CX was announced back in September 2020.
We talked about it briefly here.
However, Dialogflow CX can be pricey because it's based on the edition and the requests made during the month (a request being a call to the DF service via an API call or by using the console).
As a result, it's been nigh on impossible to really get to grips with it - the lack of a free tier/plan/trial, or a development version, has been a problem for anyone that isn't willing to pay to learn about the CX way of building chatbots.
And that's not a cool way to learn.
So, this month, along with some other important updates, Dialogflow has quietly announced a free trial version.
I say quietly because there was no mention of this in the usual release notes pages (normally, you can keep up to date with new announcements by following the Dialogflow release notes).
So, from now on, Dialogflow CX has a free trial - it's actually just a specific extension of the Google Cloud free trial.
Each new user will get $600 free credit to test and develop their CX chatbots.
We think that's great news!
What Else Was Announced?
New Dialogflow Messenger Integration
For us, this is a really important feature.
We love Dialogflow Messenger on ES, and we use this on a number of our chatbots deployed to websites.
Up until now, this has been missing which has been fairly restrictive. You can read more about Dialogflow and Messenger here.
CX Test Cases Launched
We think this is a really important feature and something that is vital once an agent is in production.
You can use the built-in test feature to uncover bugs and prevent regressions. You can read more about the CX test cases here.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Fred Whitton Challenge Chatbot Goes Live
The Fred Whitton Challenge consists of a 112 mile charity sportive around the Lake District. As the name suggests, it is run in honour of Fred Whitton. Fred, who died aged just 50 in 1998, was an extremely popular member of the Lakes Road Club.
Renowned for being one of the most difficult sportives it is an extremely popular event with over 2000 participants and is oversubscribed each year.
We are excited this year to have created the Fred Whitton Chatbot. On automated assistant chatbot to help the organisers.
It can answer questions and enquiries which come through Facebook Messenger providing important event information 24/7 and help event organisers answer enquires about the event, which has over 2000 participants.
The bot also allows users to look up past event times if they have participated previously and check weather reports and look up event facts and tips as well as vital safety information.
We enjoyed working on this project as the event is a well-known charity sportive. We had some great input from event organizers and coupled with the ability to train our bots to become smarter this has enabled us to answer a large number of typical event enquiries and successfully reduce the effort involved to manage them. The bot is currently able to answer over 90 separate enquiries.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
YouTube Adds Voice Search & Commands
YouTube uses voice search technology to augment it’s website user interface.
We covered this concept briefly in our What Can We Expect From Conversational AI In 2021 post.
One of our predictions for 2021 was the rise in the popularity of adding voice capabilities to software products. Specifically leveraging this sort of technology in touch screen situations. Software developers will improve their products by removing friction from the touch screen experience by bringing in voice controls.
Following this technique YouTube has added a voice commands input feature that can be used to search queries or navigate through videos on the streaming platform. Despite not being officially announced the feature is pretty useful for YouTube bingers.
Voice commands to search and play content
The new UI feature is simple to use and is very similar to the voice search function found on Android.
Voice commands can be used to scan the app for videos, navigate through results or pages, and even play content. The role also recognizes natural language commands for easy operations and supports multiple languages.
Using The Feature
We tested across 3 different browsers: Chrome, Firefox and Edge and had similar experiences. Click the microphone icon next to the search field. Once the user gives permission to access the microphone on the computer, a box appears with the word ‘listening’ within, and any video playing will pause. Clicking on the microphone button in the box will pause or restart YouTube’s listening for search terms.
The user says what they are looking for and then presented with the search results. Although not connected to Google Assistant the natural language seems to be fairly sophisticated.
Asking it to show you videos about "chatbot technology videos" will lead to a search of the correct term. However the natural language processing can still be tripped up with certain searches. For example it took a couple of tries to get the correct search for "rasa channel" to bring up the correct Rasa channel.
The search will also understand specific commands, for example If you give a command saying “play Rudimental" it finds and automatically plays a random song by the band. If you just say “Rudimental" in the voice command, it will open the official page of the band and display the list of their albums and songs.
The voice search feature can also search through your personal collections, listing watch histories and libraries, or gathering the latest videos from your subscribed channels. If asked to show the latest videos from a specific channel.
Using the voice search feature is not only limited to searching: it can be used to navigate to parts of the UI:
“Show me my subscriptions" will take you to your subscriptions list.
"show me my watch library" will take you to your watch library list.
Conclusion?.
The entire feature is essentially the same as the voice search feature added to YouTube’s mobile apps. In some ways, it’s surprising it's taken so long for voice commands to expand to the website although some browser limitations may have caused this.
Either way voice search is a useful feature to have and voice search for YouTube makes a lot of sense. The WhatsApp platform makes it easy to binge-watch endless similar videos and make it easy to randomly jump around. For kids who may struggle with spelling or other users with less dextrous fingers voice controls could be the new favourite tool. The feedback so far from our 10 year old tester and avid YouTube user is "that's pretty cool I will use that".
At The Bot Forge we feel that adding voice modality to user interfaces makes a lot of sense and a useful way of improving the usability of websites and software products.
Our own chatbot monitoring and analytics platform provides similar features.
Adding this type of feature could be beneficial to your website usability; get in touch, we would be happy to help!
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Google Launch AI Chatbot for COVID-19 Info
The Rapid Response Virtual Agent program includes open source templates for companies to add coronavirus content to their own chatbots.
Artificial intelligence and machine learning are continuing to take a front-row seat in fighting COVID-19, with Google Cloud launching an AI chatbot on Wednesday. The chatbot, which it calls the Rapid Response Virtual Agent program, will provide information to battle the COVID-19 pandemic, as announced in a Google blog.
The program will Google Cloud customers to respond more quickly to questions from their own customers about the coronavirus. It's designed for organizations who need to be able to provide information related to the COVID-19 pandemic to their customers, such as government agencies, healthcare and public health organizations, as well as travel, financial services and retail industries.
Google also offers Contact Center AI for 24/7 self-service support on COVID-19 questions via a chatbot or over the phone. Google also allows for businesses to add COVID-19 content to their own virtual agents with the ability to integrate open-source templates from organizations that have already launched similar initiatives. For instance, Verily partnered with Google Cloud to launch the Pathfinder virtual agent template for health systems and hospitals. It enables customers to create chat or voice bots that answer questions about COVID-19 symptoms and provide guidance from public health authorities such as the Centers for Disease Control and Prevention and World Health Organization (WHO), according to the Google blog.
The Contact Center AI's Rapid Response Virtual Agent program is available in any of the 23 languages supported by Dialogflow.
Google has provided a template to rapidly create a Dialogflow agent: You can find the template here. There is also documentation on how to build and deploy a virtual agent, whether voice or chat.
We've been looking in more detail at this template and created our own chatbot. This is a work in progress and will be something which we are updating and improving daily. You can interact with this chatbot in the bottom right of this page.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Why Should You Use Dialogflow For Your Chatbots?
Dialogflow is Google’s human-computer interaction developer which is based on natural language conversations. At The Bot Forge, Dialogflow is our platform of choice for chatbot construction.
There’s three main reasons for why we’re amongst companies such as Domino’s and Ticketmaster who make Dialogflow their chatbot platform of choice.
- Flexible coding: Thanks to Dialogflow’s in-line code editor, the time taken to complete code-related tasks is quicker than with other platforms. The prime benefit here is that we’re then able to spend more time perfecting the conversational experience.
- Scalability: Whether you start with 1,000 or 100,000 users, the platform can scale to your needs. As Dialogflow is hosted on the Google Cloud Platform, this allows the potential to support a user base of hundreds of millions, if required.
- Inbuilt machine learning: Arguably the biggest benefit of the platform in comparison to others is the availability of machine learning and natural language processing technologies. The access to these features allow us to create a richer and more natural conversational experience for your users. Dialogflow makes this possible by allowing us to extract data from a given conversation, in order to train our agents to understand user intents. Plus, as the technologies are already built into the platform, we’re able to construct your application much faster.
To ensure that we’re using the right platform for our clients’ needs, we continuously refresh our knowledge of other bot construction tools, such as The Microsoft Bot Framework. A benefit of using this platform from a developer’s perspective is the availability of templates to choose from, which allow for a more time efficient development. The IBM Watson Assistant is another platform that a developer may favour, as the testing the bot is simpler than it is on other competing platforms. If a priority is to feature your bot over a wide range of locations, Recast.AI may be a good option for its availability on 14 different platforms.
But, these platforms aren’t without their weaknesses. Unlike Dialogflow, Microsoft Bot Framework is lacking in the tools which help to create the “brains" of the bot, which is important for the sophistication that users are beginning to expect. Also, a downside of IBM Watson Assistant is the unintuitive relationship between intents (representation of user’s meaning) and entities (expressions recognised in categories). If you’re interested in how Dialogflow utilises intents and entities, we will be covering this in a future blog post.
Although we understand that there are features of other platforms which can make the development process more efficient, the inbuilt machine learning features of Dialogflow means we can deliver a bot that can produce a much richer conversational experience.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Why Business Chatbots Give You a Competitive Edge
Chatbots For Business
The business landscape is evolving faster and faster, we look at using chatbots for business to help you remain competitive.
There is so much coverage of artificial intelligence technology and chatbots these days There is no doubt that chatbots are big news for many different industries, from e-commerce and fashion to healthcare and banking.
Whilst many big brands have already jumped at the opportunity to leverage the technology for others it’s challenging to see where they can be a benefit to your company and whether the cost and effort involved is worthwhile. Some of you may remember many years ago when you were approached by someone selling a shiny website and then later on a new app? You probably asked yourself a similar set of questions..that’s nice and shiny but why do I want it? is it right for us? First things first let’s look at some of the basics.
The chatbot lowdown
What is a chatbot?
Briefly put a chatbot is a service, powered by natural language processing rules and artificial intelligence (AI), that you interact with via a voice or text-based chat interface. AI technology is used to enable the service to respond to specific user interaction. For example, a user could ask a chatbot a question or give it an instruction and the bot could respond or perform an action as appropriate.
This chat service can take on any number of roles, providing answers, collecting customer information, suggesting products and making sales. They can live in any major chat product (Facebook Messenger, Skype, SMS, Slack, Telegram, Viber, Twitter, Website). They can also be deployed into voice-enabled assistants such as Amazon Echo or Google home. Chatbots can also be developed to include multiple language capability.
Where can a chatbot be used?
Chatbots have been deployed in many different guises as they are extremely flexible and able to take on whatever business need arises.
You could say the possibilities are endless, here are some examples:
Celebrity
www.m.me/katyperry
Katy Perry’s official Facebook Messenger bot.
Customer Service
Vodafone TOBi
Vodafone’s customer service chatbot is based on IBM’s Watson & provides a fully integrated webchat for customers.
Productivity
AceBot
https://slack.com/apps/A0GRU84TF-ace
AceBot a productivity tool with expense tracking & intelligent task management, deployed in Slack.
Sports and Events
www.m.me/fredwhittonchallenge
The Saddleback Fred Whitton Challenge sportive bot is a smart events assistant providing event info to participants.
E-Commerce
www.m.me/LEGO
The official Lego Facebook Messenger bot. Ready to help your next LEGO purchase.
The benefits of using chatbots for your business
Provide stellar customer service 24/7
For many businesses, the biggest challenge to serving your customers in several communication channels is responding quickly all of the time.
Constantly available
One of the great benefits of a chatbot is the constant availability. Customer expectations are high expecting a quick response to enquiries. With a chatbot, you can offer your customers a service which is available 24 hours a day even when there are no employees in the office. You can rely on your bot no matter what time of the day or day of the week or timezone the enquiry is coming from.
One example from my own personal experience was with a SAAS which had charged me incorrectly for an amount of money which caused my bank account to go overdrawn. I contacted the customer support chatbot via a web interface at 1 AM and the problem was rectified and money returned promptly the next day. I went from disgruntled to a satisfied customer in a 5-minute chatbot interaction, incidentally, I’m still a customer!
It’s also worth noting that chatbots can be enabled to understand multiple languages. NLP technology will understand queries in different languages and respond appropriately. So if you support a global customer base needing to support multiple language enquiries this does not have to be a problem.
System integration
With the correct integration development, a chatbot is able to answer complex enquiries by integrating with existing CRM, ERP, CMS, and other business-critical applications.
Connect your chatbot seamlessly with your entire business ecosystem.
Scalable
Chatbots are scalable and capable of handling multiple enquiries, ready to step up when enquiry demands are at their peak.
A well implemented and executed chatbot can give businesses the ability to have more conversations and help more people at once than other alternatives, for example, live chat applications on websites.
This ability to handle the frequent enquires where the responses are often similar facilitates businesses in freeing up staff to deal with the more complex issues.
Although a chatbot cannot handle all customer queries, it can be used to deal with a large number of the routine business enquiries which most companies deal with on a day to day basis.
They improve customer satisfaction
To avoid frustration, a chatbot can be developed to use a “sentiment" function to pass users onto a real advisor if the bot can’t help or if they are not satisfied.
Other benefits can be seen in customer service gains. According to Jon Davies, head of digital at Vodafone, their customer service chatbot, TOBI provides “a far more engaging and personal" customer experience, as well as improving completion rates and reducing transaction times. These types of successes are highlighted in improved net promoter scores (NPS).
Overall chatbots for business can excel in supporting customer service teams in their communications with customers. Providing accessible information 24/7 saves businesses money and time. By 2022 chatbots are expected to save $8 billion.
Drive sales, engagement, reach
These days customers are savvier and demand an intuitive and seamless customer experience. Businesses need to consider using technology to fit in with their communication habits.
Familiar messaging technology
Many users prefer social media and mobile platforms for communication and expect businesses to be online when they are. If users are having a conversation with a chatbot in Facebook Messenger, they are using a conversation channel they are familiar with and they are already using the technology and don’t need to install a new app. The numbers of messenger app users have been steadily rising. As of April 2017 Facebook Messenger had 1.2 billion monthly active users worldwide
Use these channels to reach new and existing customers.
It’s also important to note that 2 out of 3 customers actually prefer to message a business to submit an enquiry rather than use other more traditional channels such as email or phone. Every day 1.4 billion people around the world send over 50 billion messages to communicate with each other. As messaging becomes even more central in people’s lives, demand for service in messaging has continued to rise.
The rise of voice assistants
Voice assistant technology and it’s adoption has gathered serious momentum over the past couple of years. User expectations are rising as they become educated in what it can do. As customers realise that its capabilities go beyond setting a timer, turning down the lights or playing some music; they will look to this channel to make purchases, contact customer support or use as a tool for business specific tasks.
The latest from Google Popular voice assistants currently include Apple’s Siri, Amazon’s Alexa, Google Now, Google Assistant and Microsoft’s Cortana. The big players are investing heavily in perfecting voice interfaces.
The reach of this sort of technology cannot be underestimated. You can read some of the stats and predictions for voice technology here.
Marketing clout
As an effective marketing tool chatbots can give your company an edge as they can enter into personalized and automated communication with your customers.
Using platforms such as Facebook messenger, substituting emails with push notifications can obtain much higher click-through rates. Used wisely opt-in targeted messages or push notifications have 90% read rates and a 40% click-through rate. Chatbots can be used to send users personalised tips, greetings and information, generating leads, harvesting reviews and forging stronger customer relationships.
Utilising these techniques a chatbot is able to reach participants wherever they are, regardless of where the chat session was initiated, whether on a mobile app, a website and even from social platforms such as Facebook Messenger.
Businesses are finding chatbots to be a great tool to engage with their market: “Our target customers are early adopters of social innovation so a chatbot is the perfect vehicle for us to communicate with them", Sarah Gower, Adidas.
Sales
Chatbots are ideal to answer first customer questions. if the chatbot decides that it can not effectively serve the customer, it can pass those customers to human agents. High value, responsive leads will be called by live agents increasing sales effectiveness.
Chatbots can be used to answer customer’s questions and promote products. Engage with the right customer by analyzing their profile and historical data and user characteristics. A bot can provide a channel for purchasing easily and quickly if requested.
Conclusion
I’ve really only scratched the surface of chatbot and voice interface technology capabilities and what can be achieved and how it can help your business be more competitive. However, it’s important to consider them carefully. It’s up to the business to decide if a chatbot is a right move for them, for some the business case may not be there or something to consider in the future. Building a chatbot because you think you should or because its the latest thing can only result in wasted time, money and effort.
I hope you find this post helpful in considering how using chatbots for business can help you to achieve a competitive edge.
If you already have a chatbot idea and want to look into this further have a look at our post planning the best chatbot
At the Bot Forge, we specialise in building chatbots.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Learn How AI-Powered Sports Software Helps Event Organisers
A Sports Software Chatbot Case Study: The Fred Whitton Challenge Sportive automated assistant, advanced
We report on our AI chatbot sports software project to aid the organisers of one of the UK’s most well-known cycling events.
Leverageing ai powered sports software with our core Aktivebot chatbot the goal was to create an automated assistant available 24/7 to reduce time and effort needed by event organisers to respond to event enquiries whilst still providing an easy way to contact the events team if necessary.
The Saddleback Fred Whitton Challenge is a charity event in honour of the late Fred Whitton consisting of a 112-mile charity sportive around the Lake District and is arguably one of the UK's most well known and hardest sportives with over 2000 riders and 5000 applications this year.
The Fred Whitton Challenge has been running since 1999 and as a result is extremely popular with over 4000 followers on their Facebook page where a large number of ride questions were being asked via the message me button there. We wanted the AI chatbot to assist the event organisers in answering ride and registration queries and reduce the amount of time spent answering routine questions. We also wanted to provide the ability for users to look up their time for this year and previous years.
The chatbot we created is integrated within the “Facebook Messenger app" of the Fred Whitton page and users can contact it through the private “Messages" feature of their page, or directly through the Messenger App.
The sports software project
The project brief was for The Bot Forge to create an AI powered chatbot capable of handling event enquiries 24/7 which could be deployed into the Facebook Messenger framework and utilise rich ui elements. Future deployments could be aimed at website integration.
For such a long-running event, Human Race and the Fred Whitton organisers wanted to provide the optimum user experience and still make it easy for participants to message organisers directly through the chatbot if they wanted to contact a real person by messaging them directly.
The chatbot understands human language, leveraging advanced Natural Language Processing and answers questions such as “what is the fred whitton?", “ I’ve injured myself at the weekend I need to defer till next year",“ when can I get my race pack?", “ help I need the GPS files for the route", “ Is there any way to buy a jersey post-event?","I want to contact an organiser", and “when will the results be available?" The chatbot replies to a question based on it’s own programmed data or points to the specific information on the Fred Whitton Website so that it works in tandem with the website itself.
[av_video src='https://www.youtube.com/watch?v=LUSPZnmiACI' format='16-9' width='16' height='9' av_uid='av-2b5n7a']
Press the play button to watch a real conversation with The Fred Whitton Chatbot
The technology
We used Google Dialogflow to provide the NLP engine and Google Firebase for the fulfilment hosting. The fulfilment or web-hook is where we were able to compute more complex answers for the AI chatbot to give to users and create the correct responses for. For example when looking up users past ride times, the web-hook was able to look up past results for users from a results database. Facebook ui elements added rich content, particularly useful when asked about merchandise details and availability; linking directly through to the official shop.
The conversations
The real challenge in creating the chatbot was leveraging natural language technology that can support the range of questions that event participants might ask: for example, all the different ways that people might ask about the route. We are helped in this process by our own Aktivebot pre-created sports events intents.
Small talk
The chatbot includes the ability to provide small talk, which is used to provide responses to casual conversation. This feature greatly improved user experience when talking to the agent.
Initial question data
Initially, we imported the pre-created sports events intents (an intent represents a mapping between what a user says and what action should be taken by the chatbot).
We then looked at FAQ data provided by the Fred Whitton steering committee and historical questions to their facebook page which gave us some invaluable insight. Using this information we were able to create the conversational scripts and then implement the conversation ability with each question matching an intent
This was an iterative process. Matching user intents to core functionality and features and training the natural language processor to understand users and handle conversation failure scenarios gracefully.
The conversational UI was then fine-tuned, with rich elements implemented where necessary.
What were the questions?
Most asked questions by participants match the questions that the event chatbot is able to answer, i.e.:
- Questions about registration: deferring places, available places, waiting list enquiries.
- Questions regarding merchandise: jerseys for sale on the day.
- Questions about the ride: route details, information about closed roads, clothing enquiries.
- Questions after the event: results, photos availability, the next ride date.
The training
The questions were often related to ride specific information. This meant that for an optimal intent matching rate, it was necessary to work closely with the event organisers to provide answers to specific questions. The capabilities of an ai sports software chatbot will improve over time, the more messaging transcript data the better so the more it’s used the better and more accurate it will get. Hence the training logs were checked multiple times a day and improvements made where necessary. By focusing on all questions answered it is possible to greatly improve the intent matching rate of the chatbot over time.
The training data was invaluable for perfecting the bot conversations. The process highlights any need for new responses as a continuous cycle of continuous learning.
The “training" of the chatbot can then be used from one year to the next. Any event detail changes can be carried out easily.
Results
The sports software chatbot was launched on 21st March with the scope constrained to Facebook Messenger with no advertising whilst the chatbot was evaluated.
Activity
The high number of participants using the chatbot can be explained by the fact that visitors still have questions that the website itself does not answer or does not answer quickly enough. The chatbot was, therefore, a great place to provide up to the minute event information, such as information about closed roads and the slight route change which resulted in one more hill showing.
The chatbot was not heavily advertised so we envisage activity levels will improve as participants get used to the chatbot as a resource they can use and other strategies to engage users are utilised.
The chatbot was answering questions on the run-up to the event and also during and after.
Success rate
The success rate of the chatbot to answer queries was overall around 60%. With more focused training over a longer period with another event in 2019 we expect this figure to rise until our aim of an 80% success rate is reached.
Feedback
The chatbot worked well in Facebook Messenger as its one of the preferred channels for chatbots in general. Deploying the chatbot in a chat widget as part of the website itself would undoubtedly result in more engagement and something to consider for the future.
Help intents and the handover protocol were also very successful. If a user did not get a correct response and/or wanted to get help or contact an organiser directly this worked really well. The overall feedback from users was positive. There were always some intents which the bot would struggle to match the first time which would be handled gracefully; however, due to the ability to train the chatbot, leveraging AI the correct response would be prepared for next time.
I’m impressed with the chatbot it seemed to work well. I think it is a good source of help and with it learning as it goes along it would answer lots of questions going forward. If it cannot help it still contacts the organisers where we can answer.
Carolyn Brown: Fred Whitton Challenge Steering Group — Saddleback Fred Whitton Challenge
The Fred Whitton Challenge chatbot still has many areas where it can be developed and improved, particularly by providing more integration with existing systems and utilising push notifications: this will be something carried out in the future.
Overall the success of the chatbot hightlights the benefits of deploying this type of ai sports software in sporting events and is definitely something to consider to give event organisers an advantage in a competitive market
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
The Bot Forge Featured in Milton Keynes Business Magazine
INBUSINESS COVERS OUR AI CHATBOT IN SCIENCE AND TECHNOLOGY SPOTLIGHT
It was great to have our AI Chatbot featured in the inBusiness magazine issue spotlight this month. You can read the feature here
Inbusiness is a bi-monthly publication and digital magazine created by distributed to over 3,000 business contacts in and around Milton Keynes. The June/July 2018 issue spotlight was science and technology so it was great that the editors of the magazine wanted to cover our Fred Whitton Challenge ai chatbot, particularly when the ai chatbot was created to assist the organisers of a charity ride.
You can learn more about our chatbot agency here.
We also cover further technical details about the project here.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Key Features of Conversational AI Platforms
We build a lot of different types of chatbots at The Bot Forge and deliver these to a variety of channels such as websites, Facebook Messenger, Slack, and WhatsApp.
To create our chatbots we often use different AI platforms which offer more suitable features for a specific project.
All the major cloud and open-source providers have adopted similar sets of features for their conversational AI platforms and provide good NLU (Natural Language Understanding).
There are also some strong options for open source privately hosted systems.
Conversational AI Platform Key Features
We wanted to spend some time looking at some of the more popular AI platforms in a bit more depth in this series. To help look at each one we have focused on the following specific features:
API & UI
A conversational AI platform should provide User Interface(UI) tools to plan conversational flow and help train and update the system
Context
As well as intent and entities, a context object allows the system to keep track of context discussed within the conversation, other information about the user's situation, and where the conversation is up to. This is often the NLP feature that is vital in creating a complex conversation beyond a simple FAQ bot.
Conversation Flow
Looking at the current position of a conversation, the context and the user's last utterance with intents and entities all come together as rules to manage the conversational flow.
This can be challenging to create and manage so a platforms' tools in the form of a flow engine, in code and complimented by a visual tool can provide advantages depending on the chatbot project itself.
Other features such as slot-filling (ensuring that all the entities for intent are present, and prompting the user for any that are missing) can be important.
Whilst most platforms fall into this category some systems use machine learning to learn from test conversational data and then create a probabilistic model to control flow. These systems rely on large datasets.
Pre-Built Channel Integrations
Having a conversational platform that supports your target channel out-of-the-box can substantially speed up the delivery of a chatbot solution and your flexibility in using the same conversational engine for a different integration.
This is one of the reasons we really like Dialogflow's tooling.
Chatbot Content Types
Whist the focus of a conversational AI platform is understanding pure text, messaging systems and web interfaces often involve other content, such as buttons, images, emojis, URLs and voice input/output.
The ability of a platform to support these features is important to create a rich user experience and help to manage the conversational flow.
Integrations
Bot responses can be enhanced by integrating information from the user with information from internal or external web services.
We use this type of ability a lot in creating our chatbots and in our opinion feel it's one of the most powerful features of a chatbot solution.
With this in mind, the ability to configure calls to external services from within a conversation and use responses to manage conversational flow is important in building chatbot conversations.
Pre Trained Intents & Entities
Instead of creating entity types such as dates, places or currencies for each project some systems provide these pre-trained to deal with complex variations.
In the same way, common user intents and utterances such as small-talk are offered pre-trained from some platforms.
Analytics & Logs
The key to creating a successful chatbot is that they need to be constantly trained and monitored.
To aid in continuously improving the system once initially launched, the conversational tools should provide a dashboard of the user conversations; showing stats for responses, user interactions and other metrics. Export of these logs is also useful to import into other systems.
Other important AI features enable easily training missed intents, catching bad sentiments and monitoring flow.
Tech Stack
It can be important to take into account what libraries are provided by an AI platform and in what supported languages.
In the end, the stack may favour a particular platform if it fitted with your current codebase or teams skillset.
However, as a full-stack javascript software house, we find Node.JS to be our server stack of choice when building our bots and most AI platforms cater for this.
Costs
These are the costs for cloud hosting and cloud NLU solutions. An important aspect to consider particularly for large scale enterprise chatbots handling large volumes of traffic where NLU monthly costs can reach £thousands.
Many providers offer a free tier for their AI platform solutions. A paid-for tier will then normally offer enhanced versions of the service with enterprise-focused features and support for greater volume and performance.
Costs tend to be charged in one of 3 ways, per API call, per conversation or daily active user and also per active monthly user (normally subscriptions are in tiers).
We try and look at costs as publicly published for the paid-for plans suitable for enterprise use in a shared public cloud environment.
The Platforms
Keeping all these feature sets in mind we hope to look at the following AI platforms over the coming posts.
- Botkit
- Chatfuel
- Amazon Lex
- Microsoft Luis
- Google Dialogflow
- Rasa
- IBM Watson
Please get in touch if you feel we should look at a platform that we have missed!
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Dialogflow Enterprise Edition Announced
It was exciting to hear that Google has announced the beta launch of its enterprise edition of Dialogflow, its tool for building chatbots and other conversational applications.
What Does This Mean For Our Customers?
It comes with a number of benefits, including:
- Basic analytics and monitoring capabilities
- Built-in support for speech recognition
- 24/7 support
- SLAs and enterprise-level terms of service promising data protection
- Higher text query quota
- Now part of Google Cloud
The new enterprise launch is an important addition, meaning The Bot Forge can provide an improved service to customers looking for an enterprise-grade chatbot.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Building a Chatbot Using Amazon Lex
What is Amazon Lex
Amazon Lex is a service by AWS for building conversational interfaces into any application using voice and text. Lex has quickly become popular among chatbots enthusiasts looking to leverage the technology which powers Alexa.
Users can be up and running within minutes with no upfront costs.
Amazon Lex has been in development since 2010 but was first made available in 2016. The system consists of the technology that powers the automatic speech recognition (ASR) and natural language understanding (NUR) capabilities of the Alexa console.
Amazon Lex is a fantastic choice for building chatbots. The main selling points are the system's ease of use, pay-as-you-go pricing, and excellent opportunities for integration over the cloud.
Artificial intelligence capabilities
Amazon Lex is powered by the same engine as Amazon Alexa. As a result, the system ranks highly in terms of speech recognition, and complex nuances, and sophisticated language understanding.
Developers can use Amazon Lex to create chatbots for voice and text which can be employed for a variety of uses, including customer service, taking orders/reservations, or controlling IoT devices.
Amazon Lex can begin building conversational interfaces with just a few example phrases. With intent chaining, the developed chatbot can suggest the next topic and switch dynamically.
Lex also supports advanced features such as Slot filling and you can meet pretty much any integration requirements by leveraging Lamba functions.
Admin platform
You must first have an Amazon Web Services account before using Amazon Lex. Access is found through the AWS Management Console.
Users can then use the Amazon Lex console to create and deploy speech or text chatbots directly to new or existing chat applications, web apps, and mobile devices like Slack, Kik, Facebook Messenger, or Twilio SMS.
Fully managed and scalable
One of the best things about Amazon Lex is that it is fully managed, so as your user base grows you don't have to worry about hardware or infrastructure. The system also offers an unparalleled opportunity for integration with other services.
Through AWS Lambda, Amazon Lex supports enterprise integrations. With Lambda, you can run code for virtually any type of application or backend service.
So integrating with CRM, ERP, Appointment systems, or content management systems is all achievable whatever the use-case.
Pricing
Amazon Lex has no minimum fees or upfront costs. Users are charged on a text or speech request basis, so you'll only be charged as much as you use it.
Lex is also a pay as you go service, so there is no recurring fees or subscription. You can also test the system and build your initial chatbot at no cost whatsoever.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
6 Common Mistakes to Avoid When Developing a Chatbot
We are experts in developing chatbots so we know If you are looking to streamline certain operations of your business, developing a chatbot is a great way to go about it.
After all, you can use technologies such as artificial intelligence (AI) and natural language processing (NLP) so that it can follow different types of conversations with users and provide relevant responses where necessary.
Chatbots are no longer a gimmicky tool available on the internet, as they have gained popularity and sophistication many users are incorporating them into their digital strategy. For example, a large number of businesses now use bots as part of their customer service. As these chatbots never go offline, they are always available to assist users, at any given time of day.
It is a lot cheaper for a business to implement a bot as part of its customer service than hire employees. If you employ an individual, you have to train the person and provide a salary and vacation time. One example is one of our clients https://amicable.io who replaced call centre resource with a chatbot to book client meetings.
Chatbots have a lot of advantages, which explains why businesses want to make the most of them. However, while creating these bots, it is natural to make errors, which hampers the user experience. As this might be the first or nth time you are developing a chatbot, you want to make sure it functions as expected. Here are six common mistakes to avoid along with how you can overcome them:
1. Assuming every user wants to talk
You tend to believe that everyone who visits your page or installs the chatbot on Slack, or other popular messaging platforms wish to start talking to it immediately. However, most of the people on the internet don’t want to communicate with the bot, unless it is necessary.
One reason why chatbots are great marketing tools is that they can engage with prospects by answering important questions. As a result, it brings down the sales friction, making it simpler for the user to invest in what you have to offer.
If your chatbot starts to message the individual as soon as he/she opens it, there is a high chance the person will find it annoying. A better practice would be to wait for the user to respond or you can leave instructions in the description on how to start conversing with the chatbot. Our sports events Facebook Messenger chatbot Carly utilises this kind of functionality enabling users to set push notifications for their any new sports events based on their own criteria.
2. Developing a chatbot but failing to track it's performance
So you've developed a chatbot but since the chatbot makes use of the latest technological advancements in the industry, you might assume that you shouldn’t keep an eye on it's performance. After all, you spent a considerable portion of your time training it, so that it can have a continuous conversation with your customers.
However, you will never know the effectiveness of your bot, if you don’t track the key performance indicators (KPI). These metrics provide a deeper insight into how you can continue to improve your chatbot. For example, you can see where most of the users tend to leave the conversation. With this data, you can think of different ways to keep them engaged so that they continue to talk to your bot. We feel that the history and training tools provided by Dialogflow enable us to track chatbot performance effectively.
3. Forgetting to list in directories
Once you have completed developing your chatbot and its up and running on various messaging platforms or your website, its a mistake to think you completed your job. All your visitors have to do is start talking to the bot, and it will help them in their tasks.
However, not everyone will know about the existence of your chatbot. Several messaging platforms may not have powerful search, which makes it harder to discover your bot. The best practice is to find third-party websites and lists your chatbot in it. As a result, if people look for your bot on Google or other search engines, the chances of it popping up in the first page of results goes up significantly. The best place to market your own new bot is on your website, why not write a blog post about your chatbot journey, you can guarantee other companies will be interested in your journey.
4. Impersonal conversations
The reason why people don’t like talking to bots is that the conversation tends to be boring and bland. As a result, they prefer to converse with human beings, as the experience is better in every way.
Think about it, would you like talking to a chatbot which sounds like it is speaking in a monotone? Rather than putting your bot in the same position, you should think of different ways to spice up the conversation. For example, you can ask the user what the chatbot should call the individual while talking to one another.
One thing is key here and we have seen this in our experience: to gain better customer satisfaction it's better to explain to your users that your bot is a chatbot and not try and masquerade as a human. If users are aware they are talking to a chatbot from the off it will gain confidence and improve the customer experience as the user becomes more forgiving.
5. Not paying attention to a chatbot's persona and tone.
Since the entire conversation between the chatbot and its users is going to take place via text, you need to pay close attention to it's tone. Deciding on a persona for your chatbot is part of the conversational design process. Using the right type of communication will determine whether your bot performs well among its intended target audience.
While this tends to be challenging, there are several ways you can overcome this obstacle. For example, you can ask a small number of people from your target audience, what tone they would find appropriate. At the same time, you can also have a beta group, which allows you to experiment and see which one works well. Matching tone to the industry and subject matter is important to build a satisfying experience for your chatbot users.
6. Help features and live agent handover
Since the purpose of the chatbot is to reduce the workload of your employees, you tend to assume that you don’t need a live agent. The problem with testing is that it may not take into account all the variables present in real-world scenarios. As a result, when you deploy your chatbot, it might not know how to handle a specific question.
Due to this reason, it can go on an endless loop, and the only way out of the conversation is to quit or restart the chatbot. An excellent way to overcome this problem is to allow your chatbot to ask a live agent to join the conversation during this situation. Once the employee helps out the user, he/she can provide information to the developers on how to improve the communication skills of the bot.
It's also important to provide easy help for users to access during the bot conversation. At The Bot Forge we always implement a help feature for our chatbots so users know what they can do and how they get back on track our Facebook Messenger chatbot for the Fred Whitton Challenge is a perfect example.
Chatbots are becoming a great tool for businesses. You can use them to make life easier for your clients, by assisting them in various functions. By knowing what the common mistakes are, you can avoid them entirely and design the best bots in the industry
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Discover The Benefits of a WhatsApp Chatbot
Creating a WhatsApp Chatbot with Twilio & Dialogflow:
In this article, I’m going to cover WhatsApp business and dive into creating a WhatsApp chatbot.
Why is WhatsApp So Important?
It's official, WhatsApp has one or two users… yes, that’s the understatement of the year!
In fact, WhatsApp has 1.5 billion users from 180 countries, which makes it the most popular instant messaging app worldwide. This messenger is handy for being secure, fast, and easy to use. It’s not just about the massive number of users though, it’s about engagement. WhatsApp users send about 65 billion messages per day, that is about 750,000 messages per second! WhatsApp usage shows no signs of slowing down.
WhatsApp for Business
In 2018 WhatsApp announced the official launch of their platform made for business.
This allows companies to communicate with clients using WhatsApp for Business instead of having to use their own personal numbers.
This will allow companies to automate, sort, and respond to messages on this incredibly popular messaging channel.
As of May 2018 WhatsApp for business had 3 million users.
Whatsapp is well known for protecting your data which includes chats, documents, status updates, photos, videos, voice messages, and calls via WhatsApp’s end-to-end encryption.
As a customer, you know you’re interacting with an officially approved business and all of your rights are protected by WhatsApp’s secure environment so its no doubt that the Whatsapp business client offering is becoming increasingly popular.
Whatsapp Business App
Business Profile
The WhatsApp business client has been built with the SME in mind. The app can help you provide customer support and deliver important notifications to your customers. These WhatsApp Business accounts help brands to improve brand loyalty. A business profile gives the company a familiar “face" and identity.
First off you need to grab the Whatsapp Business App for your mobile phone of choice which is free to download.
Users can create a business profile with helpful information for their customers such as their address, business description, email address, and website.
Steps - Update your business: Open the Whatsapp Business app → Open Settings → Open Business Setting → Business Profile.
Messaging Tools
The Business client provides some really useful automated messaging functionality.
Welcome Message
You can tailor your own greeting message and send to customers who message you for the first time or after 14 days of inactivity.
Quick Replies
Businesses can create their own standard quick reply messages to streamline their conversations and save time.
Away Message
You can tailor your own away message and reply when you are away.
Steps - Use messaging tools: Open the Whatsapp Business app → Open Settings → Open Business Setting → Select Away message/Greeting message/Quick replies.
Contact Labels
Another useful feature is the ability to organise chats and contacts with labels.
Steps - Use Labels: Open the Whatsapp Business app → Open Chat → Open Menu → Select Label chat
Statistics
The business app also provides statistics covering messages sent, delivered, read, received.
Steps - Access Statistics: Open the Whatsapp Business app → Open Settings → Open Business Setting → Statistics.
WhatsApp Business API
WhatsApp Business API is the enterprise offering for the platform.
Prerequisites
The prerequisites for using WhatsApp commercially via the WhatsApp Business API is to either apply for an own account directly from WhatsApp or to buy access from one of the official Solution Providers.
Access to the WhatsApp API has been limited, to say the least, the program is currently in a limited public preview, In fact, at the time of writing, there are only around 40+ companies listed as solution providers. You can still request access but there is no guarantee when/if this will be provided, I think Facebook will be favouring end client/solution provider applications with large estimated numbers of messages. Once you have gained access you will also have the technical challenges of getting set up. A quicker/simpler option is to use one of the solution providers for now. At least whilst you wait for your application access to be approved!
For the purposes of this article, we are going to look at using Twilio as our WhatsApp solution provider.
What's a WhatsApp Chatbot & What Are its Benefits?
A WhatsApp chatbot is similar to a Facebook Messenger Chatbot. When a user interacts with (WhatsApps) your number then the response is handled by your chatbot.
So what are the benefits?
- Customers can contact your business on their preferred platform, which they use daily
- It supports the ability to respond to customers questions right away
- You can reliably send mission-critical messages from delivery notifications to booking confirmations and delivery alerts
- Leads customers down the sales funnel by enabling them to take fast actions
- It builds trust and loyalty with customers
- Personalization of customer experience is possible by customizing the script that WhatsApp uses
- Customer communications are secure with an end to end encryption in WhatsApp
- You can send images, audio, video and pdf files via WhatsApp
Creating a WhatsApp Bike Shop Chatbot with Twilio & Dialogflow
We are going to look at building a WhatsApp chatbot for a bike shop.
The chatbot will be able to answer a simple set of bike shop related questions and book your bike in for a service.
We will use Dialogflow to create our chatbot and then connect this to the Twilio Sandbox for WhatsApp. The sandbox enables us to prototype with WhatsApp immediately, without waiting for the approval of our number.
There are some things to consider using the sandbox:
- You can only message users who have joined your sandbox. Messaging other users will fail.
- Load testing profile traffic is not supported
- The Sandbox numbers are restricted to 1 message every 3 seconds
- Sandbox numbers are branded as Twilio numbers
- You can only use pre-registered templates with the sandbox for outbound messages sent outside a WhatsApp session.
If you tell your customers that you will be using their email address and mobile phone number to send them information about your services and products, you should do that and nothing more.
Step1: Set Up Your Twilio Account
Signing up with Twilio is the next step and it's free with no need to provide a credit card, bonus!
- After opening the Twilio website, click the “Try the Sandbox today" button
- Sign up for Twilio account
- You will need to authenticate your email address.
- It's also useful to fill out the customisation clarification questions. “ Welcome! Let’s customize your experience!" However, it's not essential and you can just skip to the dashboard. We answered the questions and selected Send WhatsApp messages when asked: “ What do you want to do first?"
- Go to the Products section, then Programmable SMS and then navigate to the Twilio console
- You are then asked to activate your sandbox and agree to the WhatsApp terms of service.
- Go through the steps to set up your testing sandbox in the learn section.
- Try sending messages:
1. First send a message to the test number to link your number to the sandbox.
2. Then Send a One-Way WhatsApp Message. It's interesting to note that you must use a pre-approved template from WhatsApp.
3. Try sending Two-Way Messaging. Note 2-way messaging means you now have 24-hours between your Sandbox and your WhatsApp account, without the use of templates.
Step 2: Create your Dialogflow Agent
We won't go into detail here into how to create a Dialogflow agent you can learn more here there are plenty of good resources, we recommend taking a look at this.
If you want to use the agent we have built you can create an agent and use the restore from zip feature of the Dialogflow console to import our agent which you can download from here.
Step 3: Enable Twilio Integration in Dialogflow Agent
In the Dialogflow console → Under integrations → select Twilio (Text messaging) → in the settings window, there will be a Request URL (seen here in green).
Copy this URL and go to your Twilio account in the Sandbox configuration and paste into the “WHEN A MESSAGE COMES IN FIELD".
Once you’ve done that go back to your Dialogflow Twilio settings window and input the rest of the account details: Make sure you have your Twilio API Credentials to hand, you will need Account SID, Auth Token, Phone number — Used to authenticate REST API requests.
Steps- API Credentials: Log-into Twilio → Open Dashboard → Open Settings → General.
Step 4: Test your Agent
At this point, if you have properly carried out all the steps you should be able to send a WhatsApp message to your number and the response will come from the Dialogflow chatbot. You can see ours below. Notice the sandbox limitation; Sandbox numbers are branded as Twilio numbers.
Adding Other Cool Functionality
There are loads of other cool features we could add to our Bike Shop WhatsApp chatbot, for example:
Use the Twilio WhatsApp API to send customers a notification that their bike is ready! You can read more from the API Reference here.
The WhatsApp message can be sent using a pre-provisioned template e.g.
Hi {{1}} your bike {{2}} is completed and can be collected when it's convenient, the cost for the work is {{3}}. Details of work carried out: {{4}}
Once your Twilio number has been enabled for WhatsApp you can also create your own templates.
Going into Production?
If you want to start using the Twilio API in production you need to enable your Twilio numbers for WhatsApp.
This involves initiating a request via Twilio.
Fill out this form to send the request. Once you have provisioned your numbers you also need to provide Twilio with your Facebook Business Manager ID.
Creating a WhatsApp Chatbot with Trengo
Trengo offers an omnichannel collaborative platform for its customers.
They provide the ability to support enquiries across multi-channels.
One of the cool things about their technology is that their platform supports Business WhatsApp and they have the ability to create chatbots on their platform which can be directly linked to a Dialogflow agent, which is great if you want to create a Dialogflow powered chatbot and easily connect to a WhatsApp number!
Conclusion
WhatsApp is a platform that connects billions of users every day and is now granting businesses endless possibilities for reaching and engaging with their massive audience.
Using WhatsApp for business, companies are now able to interact with customers on the platform that they love and already use.
Hopefully, you’ve enjoyed following the article and you can see the potential for using WhatsApp chatbots in your business.
We looked primarily at Twilio as the WhatsApp API provider, however, there are other providers which we will be looking to cover in the future, here are some of them:
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
A Guide to Conversational IVR
So what is conversational IVR (Interactive Voice Response) and why should businesses care.
According to Forrester Research, customers expect easy and effective customer service that builds positive emotional connections every time they interact with a brand or organization.
For businesses improving their organization’s customer experience is a high priority.
Additionally, 40% of surveyed business leaders say that improving their organization’s customer experience (CX) is a high priority, ahead of initiatives like improving products and differentiation and reducing costs.
Despite the growth of customer service via chat and email, dialing a phone number is sometimes the best option for clients to make initial contact, resolve issues, and receive customer support. Whilst for some use-cases such as the university clearing process, it’s the norm.
Unfortunately, most outdated legacy interactive voice response (IVR) systems were never designed with CX in mind and unable to handle nuanced customer enquiries which a menu system is not designed to address. At the same time, some businesses can’t afford the staff to take phone calls 24/7 or provide enough capacity to manage spikes in demand.
What exactly is Conversational IVR?
Conversational IVR is a software system which uses voice commands from customers to allow them to interact with IVR systems over telephony channels.
The development of voice gateways has enabled modern chatbots to be connected to pre-existing telephony engagement channels via SIP/RTP and as a result massively extended the reach of chatbot technology into contact centres, enquiry numbers and helpdesks across all industries.
Voice gateways provide the technology to connect Telephony services over SIP to chatbots. For example Audiocodes PNC Voice.AI Gateway.
Bringing natural language processing(NLP) to standard telephony systems means both intent and context can be understood by these systems. It removes the burden on customers to navigate through slow, confusing and hierarchical menus and simple voice implementations of legacy IVR. And it lets customers self‑serve and resolve issues within the IVR System.
Conversational IVR takes auto-attendants and IVRs to the next level providing human like experience by enabling more human-like multi-turn interactions, leveraging natural language processing, artificial intelligence, and machine learning.
The ability to act on enquiries by extracting intent and variables from a conversation means that a conversational IVR system has the potential to work through customer enquiries, field/re-route calls effectively and enrich customer support over telephony channels.
What’s the difference between conversational IVR and legacy IVR?
The IVR systems have been around for nearly five decades now. The technology’s commercial application was rolled out in 1973 and has 1000s of present-day implementations.
In its simple guise, IVRs are a touch-tone input and voice output system. Pre-recorded messages prompted callers to put in their request by pressing keys on a phone. Unfortunately although advanced at the time, the large menus, numerous options, and using the same handset to key in inputs and listen to the voice prompts makes the experience cumbersome and not really fit for purpose when providing modern customer support.
Traditional automated call center solutions ultimately depend on number selections or similarly basic input from the customer and have minimal ability to adapt; particularly annoying when someone makes a mistake with their input such as pressing one wrong button while entering a long account number. Accessing the correct information can often be frustrating with nested menus.
Whereas traditional IVR systems had speech recognition technology to handle simple voice commands such as “yes" or “no," conversational IVR allows people to communicate their inquiries in more complete phrases via a natural language understanding. Callers can describe questions or concerns in their own words which are then matched to intent by natural language understanding.
Leveraging machine learning to improve the NLP capability, therefore, allows modern natural language processing systems to be trained to understand 1000s of different intents (questions) with each one of these intents being asked in 100s of different ways. This also means that smart IVRs can continue to improve whilst still handling failures gracefully. For example, if a caller says something which the system does not understand, it can redirect to a live agent via intelligent call routing or instead ask further qualification questions such as requesting a customer to spell out their specific details. The next time the smart IVR encounters this question(utterance) it will have learned from past experience and fully understand this enquiry.
Systems can also leverage advanced metrics such as sentiment to streamline conversations and improve customer experience with a more human-like interaction.
The best conversational IVRs enable a more free, human-like service experience for customers, who aren’t bound to a specific menu script. These smart IVRs are more capable of guiding customers toward self-service-type solutions instead of involving a live agent. Not only does this maximizes employee productivity, handle time, money saved on staffing costs, but in a best-case scenario, it makes for happier customers who get fast, personalized responses from an automated system that isn’t entirely stripped of the human touch.
Is conversational IVR right for your business?
Bigger organisations fielding large volumes of calls are likely to benefit the most from conversational IVR. However, if you experience spikes in demand and your agent’s struggle with timely answers to some of the calls they handle in peak times then a conversational IVR system may bring a ROI sooner than you expect.
Anywhere that there is repeat demand for a specific customer enquiry could mean that the intelligent use of conversational IVR could remove a business pain point, impacting call center and customer service staff positively.
Due to the flexibility of smart IVR systems, it’s often easy to create a proof of concept implementation to validate and test specific use-cases before investing fully in a system.
Voice gateway technology also enables modern IVR to connect easily with current legacy systems.
What are the benefits of using conversational IVR
Its important to understand how conversational IVR can be leveraged to cut budgets, increase efficiency, improve customer satisfaction and meet spikes in demand.
Serve your customers faster with more precision
Conversational IVR reduces the amount of time needed to support each client. It often decreases the time taken to serve customers as users are often able to request their intention with one sentence, rather than navigate through confusing menu systems.
Leveraging natural language and machine learning to improve responses therefore enables Conversational IVR systems to always be improving and still be flexible enough to be easily changed to meet new questions and offer more detail where needed.
Cut costs
Its difficult to calculate the average cost of a live agent customer service phone call due to the number of variables. However there is no denying the need. In one report, IBM reported that worldwide companies spend over $1.3 trillion to serve 265 billion customer calls each year.
In a report by Forrester¹ on the total economic impact of smart IVR technology the projected return on investment (PROI) was judged to be between 103% and 291% over 3 years. Most organisations would benefit from using conversational IVR to satisfy customers without using a human agent. Financial gains are seen from:
- Reducing the number of calls reaching human agents by improved initial resolution rates and containment.
- Improving agent efficiency with augmented technology such as agent assist and predictive analytics.
- Improving customer satisfaction and agent experience by reducing the burden on agents so they can focus on addressing each customer’s specific request or need.
¹ New Technology: The Projected Total Economic Impact™ Of Google Cloud Contact Center AI 2020
Improve customer satisfaction
It’s no surprise that historical feeling towards automated phone services was negative due to the poor customer experience from legacy IVR systems. Overall voice was often judged suspiciously. These days opinions are shifting toward acceptance amid the rising adoption of personal assistants such as Google Assistant, Siri, Cortana, and Alexa on smart device and mobile. People are increasingly familiar with these technologies, the ways they can be used, and their limitations. Customer satisfaction numbers for popular voice-controlled assistants are as high as 80% or better depending on the platform and survey.
Users realise that they can get more done with voice these days. At the same time voice technology and capabilities such as agent assist can also empower agents to provide better service. With the careful implementation and design of voice assistants it’s possible to achieve higher levels of customer care and improved efficiency and as a result provide a better service for your customers.
Conversational IVR Provides Better Customer Support
Customer satisfaction is a critical concept for customer success professionals to understand and live by, and it’s actually about more than a money-back guarantee.
Conversational IVR modernizes conventional IVR principles with innovations such as AI and machine learning. Instead of navigating push-button menu flows, users can provide spoken inquiries reducing the friction and poor user experience of legacy systems. Smart IVR has the ability to provide responses to complex enquiries and give a response in real-time within seconds. Tasks that used to require a lengthy phone call can now often be done quickly and easily.
At the same time technology can be leveraged to provide streamlined features: sentiment analysis, call routing, fall-back handling, even energy detection tracking to provide better support. And if the customer needs to speak to someone or requests to do so, we can connect them to an agent seamlessly and retain all the information already shared in the chat.
The end result is that this allows conversational IVR systems to complete requests faster via customer self-service options in real-time. Along with maximizing efficiency and helping to offset spikes in call volume, call center costs are reduced by lowering customer churn, boosting brand perception, and improving client retention.
How to Train a Chatbot
Training data for chatbots.
I'm going to look at the challenges in creating a chatbot which can answer questions about its specific domain effectively. In particular, I'm going to look at the challenges and possible solutions in creating a chatbot with a reasonable conversational ability at their initial implementation. Every chatbot project is different but often clients come to us with a large knowledge base which they want a chatbot to support from its release but with very little training data.
We are going to concentrate on a Dialogflow project to look at some examples however the challenges and solution are similar for all the most well know NLP engines, Watson, Rasa, Luis etc.
The Challenge
One of the key problems with modern chatbot generation is that they need large amounts of chatbot training data.
If you want your chatbot to understand a specific intention, you need to provide it with a large number of phrases that convey that intention. In a Dialogflow agent, these training phrases are called utterances and Dialogflow stipulate at least 10 training phrases to each intent.
Depending on the field of application for the chatbot, thousands of inquiries in a specific subject area can be required to make it ready for use with each one of these lines of enquiry needing multiple training phrases.
The training process of an ai powered chatbot means that chatbots learn from each new inquiry. The more requests a chatbot has processed, the better trained it is. The NLU(Natural Language Understanding) is continually improved, and the bot’s detection patterns are refined. Unfortunately, a large number of additional queries are necessary to optimize the bot, working towards the goal of reaching a recognition rate approaching 90-100% often means a long bedding in process of several months.
Data Scarcity
One of the main issues in today's chatbots generation is that large amounts of training information are required to match the challenges described previously. You have to give it a large number of phrases that convey your purpose if you want your chatbot to understand a specific intention.
To date, these large training corpus had to be manually generated. This can be a time-consuming job with an associated increase in the cost of the project. One of the main issues we have faced is that often clients want to see quick results in a chatbot implementation. These types of chatbot projects are often use cases which are providing information regarding a wide-ranging domain and may not necessarily have a lot of chat transcripts or emails to work with to create the initial training model. In these cases there is often not enough training data and so it takes time to get decent and accurate match rates.
The Solution
THE BOT FORGE PROVIDES
CHATBOT TRAINING DATA
CREATION SERVICES
The Bot Forge offers an artificial training data service to automate training phrase creation for your specific domain or chatbot use-case. Our process will automatically generate intent variation datasets that cover all of the different ways that users from different demographic groups might call the same intent which can be used as the base training for your chatbot.
Multi NLP platform support
Multi-language support
Our training data is not restricted solely to Dialogflow agents, the output data can be formatted for the following agent types:
- rasa: Rasa JSON format
- luis: LUIS JSON format
- witai: Wit.ai JSON format
- watson: Watson JSON format
- lex: Lex JSON format
- dialogflow: Dialogflow JSON format
We provide training datasets in 100+ languages
We offer our synthetic training data creation services to our chatbot clients. However, if you already have your own chatbot project and just want to boost its conversational ability we can provide synthetic training data to meet your needs.
Testing the Solution
We wanted to test the effectiveness of using our synthetic training data in a Dialogflow chatbot agent by varying the number of utterances per intent using our own synthetic training data.
Dialogflow test agents
We carried out three different tests (A B and C) with 3 separate Dialogflow agents. Each agent had identical agent settings. The agents had 3 identical intents to provide information about the topic of angel investors:
- what_is_an_angel_investor
- what_percentage_do_angel_investors_want
- do_angel_investors_seek_control
In the first test (A) the chatbot was trained with 2 hand-tagged training phrases (utterances) per intent. Test (B) had 10 training phrases from our own synthetic training data per intent and test (C) had between 25 and 60 training phrases per intent.
The Test
We tested each agent with 12 separate questions similar to but distinct from the ones in the training sets.
We didn't carry out any training during testing once the chatbots were created.
We recorded the % of queries matched to the correct intent, the incorrect intent or no match and also the intent detection confidence 0.0 (completely uncertain) to 1.0 (completely certain) from the agent response.
Overall test results
|
|
% correct match rate
|% incorrect match||
|
%no match
|
|
Average Intent Detection Confidence
|Test A (2x utterances)||50%||42%||8%||0.6437837225|
|Test B (10x utterances)||91%||9%||0%||0.7590197883|
|Test C (25-60x utterances)||100%||0%||0%||0.856748325|
Test A provided a 50% match rate. We observed a significant improvement in test B with the introduction of some of our synthetic training data to the agent. We were able to improve the match rate from 41% to 91% whilst TestC with 25-60 training phrases enabled a match rate of 100%. The average intent detection confidence also grew
In summary, chatbots need a decent amount of training data to provide accurate results. If there is not enough training data then a chatbots accuracy is affected and it can take some time to train it whilst being used to reach acceptable performance levels. At the same time, it can be costly and time-consuming to create training data for a chatbot needing to handle large numbers of intents.
Our synthetic training data creation service allows us to create big training sets with no effort thus reducing initial costs in chatbot creation and improving the usability of a chatbot from the initial release stages. If you only have a limited number of training phrases per intent and have large numbers of intents, our service is able to generate the rest of variants needed to go from really poor results to a chatbot with greater levels of accuracy in providing responses. We have carried out these tests with Dialogflow, but our conclusions are relevant for ML-based bot platforms in general. We can conclude that our Artificial Training Data service is able to drastically improve the results of chatbot platforms that are highly dependent on training data
Chatbot Training Never Ends!
I've looked at the benefits of using our training data at the early stages of a chatbot project. However, it's important to note that the key to success, in the long run, is to constantly monitor your chatbot and continue training to get smarter. Either by doing constant training with human effort or by scheduling regular training cycles, incorporating new utterances and conversations from real users.
If you want to know more about our chatbot training data creation services get in touch
Appendix
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
The Non-Technical Guide to Popular Conversational AI Terminology
Conversational AI Terminology Cheatsheet
Conversational AI technology is not new, but the advanced in the technology has driven a major growth in the industry and what can be achieved in its role solving business problems for many types of industries.
We talk about Conversational AI a lot on our website and blog, after all this technology is at the core of what we do at The Bot Forge.
You may well have encountered some of the different terminology used. But what do developers and technologists really mean when they use these terms? Having a simple understanding of some of the more frequently used terms can be useful when thinking and talking about your chatbot or voice assistant strategy. This conversational AI terminology cheatsheet aims to help you understand; no technical knowledge required!
-
Algorithm
An algorithm is a formula for completing a task. Wikipedia states that an algorithm “is a step-by-step procedure for calculations. Algorithms are used for calculating, automated processing and data processing and provide the foundations for artificial intelligence technology.
-
Artificial Neural Network
Artificial Neural Networks or ANN are artificial replicas of the biological networks in our brain and are a type of machine learning. Although nowhere near as powerful as our own brains they can still perform complex tasks such as playing chess, for example AlphaZero, the game playing AI created by Google.
-
Artificial Intelligence
AI research and development aims to enable computers to make decisions and solve problems. The term is actually a field of computer science and is used to describe any part of AI technology of which there are 3 main distinctions (1)
-
Big Data
Big data describes the large volume of data – both structured and unstructured – that floods through a business and its processes on a day-to-day basis. In the context of AI big data is the fuel which is processed to provide inputs for surfacing patterns and making predictions.
-
Chatbots
I think we have mentioned these once or twice! A chatbot is a conversational interface powered by AI and specifically NLP. They can be text-based, living in apps such as Facebook Messenger or their interface can use voice-enabled technology such as Amazon Alexa.
-
Cognitive
Cognitive computing mimics the way the human brain thinks by making use of machine learning techniques. As researchers move closer towards transformative artificial intelligence, cognitive will become increasingly relevant.
-
Conversational Design/Conversational Designer
Whilst not a technical term its a relatively new role which has grown to being a vital one with the rise in the popularity of conversational experiences. It's important to understand what this new breed of skilled professional brings to a chatbot project and why they are so important. Conversation design is the art of teaching computers to communicate the way humans do. It’s an area that requires knowledge of UX design, psychology, audio design, linguistics, and copywriting. All of that put together helps chatbot designers create natural conversations that guarantee a good user experience.
-
Deep Learning
Also known as a deep neural network, deep learning uses algorithms to understand data and datasets. Deep Learning is a subfield of machine learning concerned with algorithms inspired by the structure and function of the brain called artificial neural networks. Deep Learning techniques have become popular in solving traditional Natural Language Processing problems like Sentiment Analysis.
-
Entity and Entity Extraction
Entities are also sometimes referred to as slots. An entity is used for extracting parameter values from natural language inputs. Any important data you want to get from a user's request will have a corresponding entity. Entity extraction techniques are used to identify and extract different entities. This can be regex extraction, Dictionary extraction, complex pattern-based extraction or statistical extraction. For example, if asked for your favourite colour you would reply "my favourite colour is red". Dictionary extraction would be used to extract the red for the colour entity. What that means in the real world is types of product, locations, model numbers, parts numbers, courses etc: basically anything related to your business which needs to be understood and extracted from the conversation.
-
Intelligent Personal Assistants
This term is often used to describe voice-activated assistants which perform tasks for us such as Amazon Alexa, Google Assistant, Siri etc instead of text-based chatbots.
-
Intent
An intent represents a mapping between what a user says and what action should be taken by your chatbot. A good rule of thumb is to have An intent is often named after the action completed for example FindProductInformation, ReportHardWareProblem or FundraisingEnquiry.
-
Machine Learning
Machine Learning or ML for short is probably used by you every day in Google search for example or Facebooks image recognition. ML allows software packages to be more accurate in predicting an outcome without being explicitly programmed. Machine learning algorithms take input data and use statistical analysis to predict an outcome within a given range. Machine learning methods include pattern recognition, natural language processing and data mining.
-
Natural Language Processing
Natural language processing (NLP) is broadly defined as the automatic manipulation of natural language, like speech and text, by software. NLP is a branch of artificial intelligence that helps computers understand, interpret and manipulate human language. NLP draws from many disciplines, including computer science and computational linguistics to fill the gap between human communication and computer understanding.
-
Natural Language Understanding
A subfield of NLP called natural language understanding (NLU) has begun to rise in popularity because of its potential in cognitive and AI applications. NLU goes beyond the structural understanding of language to interpret intent, resolve context and word ambiguity, and even generate well-formed human language on its own.
NLU algorithms tackle the extremely complex problem of semantic interpretation. That is understanding the intended meaning of spoken or written language. Advances in NLU are enabling us create more natural conversations.
-
Sentiment Analysis.
Sentiment Analysis is the process of determining whether a piece of writing is positive, negative or neutral. More advanced analysis would look at emotional states such as "angry", "sad", and "happy".
-
Utterance
An utterance is anything the user says via text or speech. For example, if a user types “what is my favourite colour", the entire sentence is the utterance.
-
Conversational IVR
Conversational IVR is a software system which uses voice commands from customers. This allows them to interact with IVR systems over telephony channels.
Whereas traditional IVR systems had speech recognition technology to handle simple voice commands such as “yes" or “no". Conversational IVR allows people to communicate their inquiries in more complete phrases via a natural language understanding. Callers can describe questions or concerns in their own words which is then matched to an intent by natural language understanding.
We hope you have found this Conversational AI Terminology Cheat-sheet helpful.
Comment if you think I've missed any terms out which should be on the cheat sheet.
If you want to talk about your chatbot project why not book a free consultation with us.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Dialogflow Migration V2 Guide [Part 1]
***UPDATE***
Dialogflow have extended the V1 API shutdown deadline to March 31st, 2020. https://cloud.google.com/dialogflow/docs/release-notes#November_14_2019
Winter is coming! (for any Game of Thrones fans this will make perfect sense!) In October last year, we wrote about the news that Google will be dropping support for V1 of the Dialogflow REST API in Oct 2019. We've been building all our chatbots with V2 since last year, however, there are many companies who still have V1 Dialogflow agents which will need to be transferred. This blog post aims to help you with carrying out your migration successfully.
The amount of work needed will really depend on what features your Dialogflow agent is using and where it's integrated: If you are using Dialogflow's fulfillment webhook, inline editor, or any Dialogflow API, you'll need to update your code, endpoints, and/or fulfillment to be compatible with V2. However If you are certain your existing agent doesn't use the fulfillment webhook library, the Dialogflow API, or any integrations, then you will not need to make any major changes before selecting V2.
Due to authentication changes, the biggest impact will be for Dialogflow web agent implementations which are currently calling the REST API.
This post will be split this 2 sections: a basic migration guide for agents not using the REST API and a more advanced version covering what changes are needed to use the new REST API and what changes need to be made to support authentication.
You can see more details about upgrading from V1 to V2 in the official guide here.
Basic Migration
Anyone who already has built out their website chatbots using v1 API, then they should start planning for the migration sooner rather than later. Any new features should be added after the upgrade. The migration is potentially a non-trivial task, considering some chatbots have some fairly complex code driving their fulfilment. If you have a live bot in production our advice is to set up an upgrade chatbot as a copy of your existing bot project and then work through the upgrade there. You can guarantee that changing to V2 will mean that fulfilment and API calls may stop working. Once the upgrade is complete re-testing all bot functionality is strongly advised before setting live.
Chatbot Web Interfaces
We would recommend everyone who is creating custom website chatbots to do so using the v2 API. All our new chatbots are built using the v2API.
If you need assistance or advice with your own chatbot v2 upgrade please get in touch, we are Dialogflow experts and would be happy to help!
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Buckinghamshire Business Festival Sponsor 2021
We are proud to be a Buckinghamshire Business Festival sponsor this year
We are proud to be sponsoring the 2021 Buckinghamshire Business Festival, running from April 19th – 30th.
The festival has been organised by Buckinghamshire Business First, with a packed schedule of events and opportunities to make new connections across the two weeks.
Look out for our article in the Buckinghamshire Business First Sponsor Newsletter due to be sent on 8th April – we are excited to be inviting businesses to contact us to find out more about what we do.
Visit the Buckinghamshire Business Festival webpage to find out more about the Festival and to see the full schedule of events: https://bbf.uk.com/events/buckinghamshire-business-festival
Use the hashtag #BucksBizFest on social media to get more involved with the festival in the build-up and as it happens.
Contact Buckinghamshire Business First for more information on the festival: 01494 927130 / events@bbf.uk.com
Sign up for a free conversational AI strategy consultation
This event is part of the 2021 Buckinghamshire Business Festival, running from April 19th – 30th. The festival has been organised by Buckinghamshire Business First, with a packed schedule of events across the two weeks.
“With our 30 minute conversational AI strategy consultation, we aim to identify how conversational AI technology can help your organisation.
Firstly we will start by gaining some understanding of your core business. Then discuss any business or process problems or challenges which can be addressed with a chatbot or voice assistant.
We will tailor our consultation to cover the areas of conversational AI technology and benefits you are interested in and where and how they could be applied to your organisation."
Adrian Thompson
In the meantime you can read learn more about chatbots and voice technology in our blog.
Visit the Buckinghamshire Business Festival webpage for the full schedule of events: https://bbf.uk.com/events/buckinghamshire-business-festival
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
What Makes a Successful Chatbot Project?
Tips. Insight. Offers. Are You In?
Your Chatbot Project
In this article, we will show how we can carry out your successful chatbot project at The Bot forge. We’ll share a few things we’ve learned from building chatbots. We’ll look at messaging platforms, voice interfaces and the importance of conversation design. Giving you a walk-through of how your chatbot project would progress, forging the perfect chatbot.
What should my chatbot do?
Right from the beginning of the project, It's important to have a clear understanding of what your chatbot will do. We like to ask the question:
What is the number one reason my chatbot will exist?
Chatbot platform
Your audience should drive your chatbot platform choice if possible. If you can collect information on which messenger platforms your audience use then this should assist your decision. Facebook Messenger is the most popular with over 1 billion active users as well as being constantly improved by the Facebook team. This is our favourite messaging platform. We particularly like the UI elements which we can provide using the Facebook platform. Have a look at our Customer Support chatbot to see an example of these elements.
If you want your chatbot project to live in a voice interface we recommend either Amazon Alexa or Google Home. The Amazon Echo has brought the voice interface to over 20 million homes around the world. Alexa is not just for handling home tasks Alexa for business means she can help you at work, acting as an intelligent assistant and integrating with your enterprise systems.
Chatbot Requirements
We like to capture chatbot requirements as user stories. The story is in the same format:
As a , I want , so that for example:
- I'm a hotel guest, I want to book a table in the restaurant, so that I can eat tonight
- As a hotel owner, I want to collect a guest reviews, so that I can improve their experience.
- As a previous customer, I want to have similar clothing choices recommended so I can match my style.
Chatbot Persona/Personality
Decide on bot’s personality and tone of voice. This character can then be used in planning the conversation in later steps forming the bot's persona in conversing with users.
We recommend creating a complete agent character which we can then model and grow, possibly adding humor to responses. Our visual designers can assist in coming up with a chatbot character.
Chatbot Conversation Design
We prioritise the user stories and plan them our in more detail. Elaborating as we go through each one. and noting main entry points to back-end system integration.
Design the conversational workflow at a high level. Record all the possible topics and conversation parts, a whiteboard session is great at this point and a simple mind map.
The next step is the conversational ui scripts(cui). We will write the bot scripts incrementally, starting with the core functionality and then expanding into personality driven intents and multiple responses. Always trying to focus on one conversational part at a time. Some examples of the scripts:
Hi there, I'm the macbot ready to give you weather forecasts hi macbot where do you live, so I can send you weather forecasts. leeds ah ok, nice city right now its 1°C I hope you've got a hat and gloves!
As part of the script design we will also elaborate on other elements of the dialogue, for example sets of options, conditions, user input and entities from the input.
We will also define any custom entities for the chatbot. The entities are used for extracting parameter values from natural language inputs. For example the following entities:
- Total Spa Experience
- Experience Wellbeing Massage
- Experience Body Pumice
Then if you said "can I book a total spa experience" the chatbot would be able to pick out the total spa experience in your reply.
{Hi, hello, good-day} I'm the spaworld bot which spa treatment did you want to book? can I book the [body pumice experience] please sure, what time and date? [midday] on the [23rd of feb] ok sure, same as last time right, with Anton? actually is [Denise] available yes no problem, all booked
We will agree on the core conversation dialogues and fine tune them. Then we will also plan on how to handle users straying away from what we call the happy path (following the normal conversational flow). We will also allow for users trying to challenge the bot with sexts, swearing, off context questions, swearing or gibberish. We will always keep an eye on the conversation goals and insure they offer the best user experience whilst making sure the intents match core functionality.
Chatbot UI Design
The BOT Forge will produce prototypes for each core intent as an interactive mock-up providing the visual text interface and voice interactions to show how the conversation will flow.
We also decide on whether we will use different types of structured messages (images, buttons, quick replies, lists, web-views. Depending on the chosen chatbot integration the mock-ups will also include some UI elements specific to the platform. We can share these mockups online or provide them as an animated gif. This is a great way of showing how a bot will work in real chatbot environment, or how the conversation will sound in the voice interface.
Chatbot development
We use an Agile development process using sprints, releasing features little and often to meet the story features. We work closely with our clients, always testing, improving the bot flow, the conversational knowledge base, the bots personality and the overall user experience. This process of iterative delivery a working chatbot will be deployed and ready to use by real users, right from the very first sprint.
Chatbot Platform Integration
We like to carry out the integration work for the chatbot as one of the first pieces of development work. This is so we can provide a working bot for our clients to be able to see a beta/testing version of the chatbot as early as possible. Then we can release bot features and conversation intents regularly.
Conversation Development
The conversational ability will then be implemented in the chatbot following the conversation design and initially focusing on the core intents and text responses. This will be an iterative process. Matching user intents to core functionality and features and training the natural language processor to understand users and handle conversation failure scenarios gracefully. As this stage some integration features will be mocked to return dummy data. During this development stage the chatbot will be provided as a beta implementation so that its available for its first conversations with our client. Users can be notified of new intents for testing.
The training data at this stage will be invaluable for perfecting the bot conversation. This process will also highlight any need for new responses as a continuous cycle. Testing the bot based on responses will continue, we call this supervised learning .
Further UI elements will also be created dependent on the chosen integration platform for the chatbot.
Integration Development
Once the conversational ability has been implemented we will implement any integrations needed for the core chatbot functionality. Writing the code to connect and extend your backend services and integrating with external services needed for the bot to deliver the correct functionality.
Working through each story element and replacing any mocked data entry points. Each piece will be unit and system tested.
Chatbot Alpha/Beta Testing
Once development has been completed we will define how long we are going to do Alpha and Beta testing. Alpha testing is a type of acceptance testing; performed to identify all possible issues/bugs and continued supervised learning , before releasing the chatbot to your users. Bugs will be logged and tracked on our tracking tool and prioritized and fixed on a regular basis. The aim is to carry out the interactions with the bot that a typical user might perform. Making sure to carry out each user story to get the expected outcome. We are happy for our clients to become involved although to be fair you will already of had exposure to the bot as part of the ongoing agile project.
Beta Testing of a chatbot is performed by "real users" of the software application in a "real environment".
Ideally the chatbot Beta version is released to a limited number of end-users of the product to obtain feedback on the product quality. Beta testing reduces conversation and integration related failure risks and provides increased quality of the user experience through customer validation.
It is the final test before shipping the chatbot to your customers. Direct feedback from customers is a major advantage of Beta Testing. This testing helps to tests the bot in real time environment.
The experiences of the early users are passed on to the developers, who make final changes before releasing the bot commercially.
Chatbot Deployment
Once the round of beta/alpha testing has been completed we can deploy the chatbot as a live application.
Chatbot Future
Of course it doesn't end there. Once deployed the chatbot project will be maintained by The Bot Forge as a yearly subscription. We will constantly monitor your bot carrying out daily supervised learning and weekly improvements. Monitoring conversations and confirming qualified intents as well as checking for unmatched intents and fixing them as needed.
We will carry out third party and integration maintenance, monitor api's for version updates. Making sure your bot is performing well and healthy!
The Bot Forge are always available to discuss further improvements and functionality to add to your chatbot or just to talk to us about your next great idea.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
10 Questions To Ask When Planning a Chatbot Project
Tips. Insight. Offers. Are You In?
Planning The Best Chatbot
Congratulations you've had that lightbulb moment and you have an idea to create the best chatbot, or maybe you've heard so much about chatbots lately you feel you should explore the idea of using one for your company. If you want to read more about how a chatbot can help your company read our Why using chatbots for business can help you remain competitive blog post
Whatever the reason now it's time to start looking at your idea in more detail to plan the best chatbot..but wait that can be hard. Don't worry, we can help!
Chatbots have progressed rapidly over the past couple of years with advancements in Natural Language Processing (NLP) utilized used across voice and text-driven interfaces. There has never been a better time to start a conversational UI project. However its still vitally important to plan your project carefully.
At The Bot Forge, we like to ask our clients the following questions to ensure we have a clear understanding of what they want to achieve with their chatbot project.
So whether you are looking at creating your own bot, or commissioning a team of chatbot experts like ourselves ;) then it's important to ask yourself the following 10 questions before you start building the best chatbot.
1. What is the purpose of your chatbot?
Why do you want to create a chatbot? What do you want the chatbot to do for your business and how will it achieve your business goals?
Right from the beginning of the project, it? s important for yourself and your team to have a clear understanding of what your chatbot will do
2. What are the key goals of your bot project?
What are the main aims of your conversational ai project? It could be to drive sales, provide 24/7 customer support or engage with new and existing customers by gathering customer feedback and delivering new product information.
3. How will you measure your success
How will you determine the success of your chatbot? What will your Key Performance Indicators (KPI) be?
For example, you could look at click-through rates, the numbers of inquiries handled correctly or feedback statistics gathered.
4. Who is going to use your chatbot?
Have a clear idea about who is going to use your chatbot, what will be the user demographic? This may influence your chatbot's persona.
5. Where will your chatbot be deployed?
Your audience should drive your chatbot platform choice if possible. You can deploy conversational ai assistants to a lot of places:
- Facebook Messenger Chatbot
- WhatsApp Messenger Chatbot
- Telegram Chatbot
- IBM Watson Chatbot
- Slack Chatbot
- Twilio
- Microsoft Teams Chatbot
- Custom Website Chatbot
- IoT
It really all depends on your use case. If it's an internal tool for your HR team then Microsoft Teams makes sense. Or if you need to help your website users then you can create a website chatbot.
If you can collect information on which platforms your audience use then this can assist your decision. Now is also a good time to consider voice platforms such as Alexa or Google Home.
6. What will your chatbot do?
Here you can really start to consider what sort of functionality the chatbot needs to provide and most importantly the conversations it will be able to support.
A good way to capture chatbot requirements is by looking at them as user stories. The story is in the same format: As a , I want , so that: for example:
- I'm a participant, I want to check what time I can start my event, so that I can be ready to leave in good time.
- As a business, I want to collect customer reviews, so that I can improve their experience.
- As a customer, I want to access my account details quickly and receive an account update through my personal assistant.
7. Will the chatbot have a character?
Will the chatbot have its own persona, will it have a character?
Is the chatbot going to just be a polite assistant or does it need a character to carry through your brand?
8. How will the chatbot create value?
Think about the overall user experience. How will the chatbot ensure that users come back?
For example by providing a simple and well-executed personal assistant then customers are going to use this as their first port of call to find information and/or contact your company.
9. How will people find the chatbot?
How are you going to drive people to find and use your chatbot?
Links on your website and also advertising on Facebook can be great places to start as well as content on your Facebook page.
10. How will you look after your chatbot?
How will you monitor performance after launch? In comparison to other projects, it's important to note that once the chatbot is launched this is just the start of your journey. Essentially you are at the start of the optimization phase. You will need to provide resources to get the most out of your automated assistant after it's gone live.
You will need to monitor user interactions, reactions, unanswered requests: so you can train and improve the overall user experience, training your chatbot is key!
Conclusion
After working through these 10 questions you should be well on your way to understanding your chatbot concept.
With all our new clients at The Bot Forge chatbot agency, we ask them to fill out our chatbot checklist, feel free to download and work through it with other members of your team.
We hope you find this post helpful in getting to grips with your chatbot project, feel free to share if you find the 10 questions useful.
At The Bot Forge, we specialize in conversational AI so why not book a free consultation with us.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
6 Tips to Ensure Your Chatbot is GDPR Compliant
Tips. Insight. Offers. Are You In?
General Data Protection Regulation (GDPR) entered into force and was fully operational as of May 25th 2018. You can read all about it here. The new regulations brought a series of changes and improvements while strengthening the current regulatory framework. The GDPR applies to any website or mobile application collecting data from EU residents and that means chatbots and voice assistants as well!
Despite some myths and misunderstandings around GDPR the regulations there has been some success in the new policy despite still being described as being in a transition period. With incidents such as the Cambridge Analytica scandal last year users are even more concerned as to what we do with their data.
It's important to note that, 71% of UK adults want tougher action in penalising companies that abuse our data privacy by misusing third-party data.
If you use chatbots as part of your sales and marketing strategies, you’ll need to make sure the processes you use to collect consumers’ personal data, as well as what you do with this data are in line with GDPR. Read on for some tips on how to ensure that your chatbots are GDPR compliant.
1. User Consent
Consent is not valid unless it is “freely given, specific, informed, and unambiguous." Basically, that means a “clicked" agreement is required.
For websites, your privacy notice is a great place to get consent from users. Here is a great example:
Don’t forget to update your privacy policy!
One of the rules of the GDPR is that all companies utilizing consumer data need to have a clearly stated privacy policy which contains the following pertinent information:
- What information is collected?
- Who is collecting it?
- Why is it being collected?
- How long will it be used for?
- Who will it be shared with?
- How can consumers withdraw from the agreement to give their data?
For a chatbot, it should provide users with a clear-cut, transparent, distinguishable, and easily accessible form to understand what data is collected, and how it will be used by the bot and organization. This needs to be provided at the start of the conversation and also its often a good idea to provide an easy way to access this in future e.g for bots supporting NLP a free text intent or part of an integration menu such as Facebook Messengers:
We've found that having a privacy page in place listing all the important information is also an effective way to aid in compliance.
2. Allow users to have their data forgotten
According to the GDPR, users should be able to request that all their Personal Data is removed.
Chatbots need an intent to support this e.g ‘please forget my data’, ‘delete my personal data’, etc. Or this could be part of the menu system:
This data removal request needs to be followed up correctly.
3. Allow users to retrieve their data
Users should be able to retrieve their Personal Data.
Chatbot users should be provided with a clear and simple way to access, review and download copies of their data (in an electronic form) that was collected, free of charge. This can be actioned in multiple ways. You could either build a dialogue for this e.g ‘please tell me what data you are storing’, ‘can you send me my data’. The response should present the data to the user or send an email to start the process.
4. Use personal data for the stated purposes only
This is vital for becoming GDPR compliant. Your online chatbot may be an informal way of collecting personal data, but it is still considered to be a data collecting and processing tool and so will fall under the GDPR legislation.
Clearly stating what information is used for is key. This means that you are only able to use the data for the stated purposes, such as sending newsletters, emails, SMS marketing messages or contacting users on Facebook Messenger.
Implement a mechanism to make sure users are clear as to what you will do with their data. This can be added as part of a welcome or supported by intent match or part of the privacy policy.
If you tell your customers that you will be using their email address and mobile phone number to send them information about your services and products, you should do that and nothing more.
5. Leverage Chatbot Conversation
Chatbots provide an engaging interaction medium for users which is no doubt enhanced by a personalised experience. This will often mean that a chatbot needs to collect some personal data from their users. When designing chatbots always remember to keep privacy first in mind. With a chatbot, it is easy to ask for a users permission and explain why you need it because you are already in a dialogue with your user.
Use opportunities when available to clarify and advise users during the conversation.
6. Safeguarding Data
Roles
There are two important roles defined in the GDPR that affect you as a company and the chatbot you build. Firstly, the data controller and secondly, the data processor:
- Data Controller represents the entity which determines the purposes and means of the processing of personal data
- Data Processor represents the entity which processes personal data on behalf of the controller
Data controllers are the decision makers about which personal data gets collected, stored and processed - so most companies are considered controllers!
Chatbots are all about data. If you want to create a solid conversational experience, you need to use Natural Language Understanding (NLU) and dialogue systems. The underlying machine learning algorithms need training data in order to improve and learn. Collecting this data is necessary to train the models and the more data you have the better the bot performs.
Data is essential - but it's also vital to reduce the risk of data breaches and adhere to the GDPR data processing principles.
With GDPR you are prohibited to store this data without explicit consent from users or if there is no legitimate reason to store this data. If you do have a need to store this data to improve your chatbot’s interaction with consumers, you may not do so unless you have explicit consent.
It’s common for many web and messenger servers to keep different types of logs, such as access, error or security audit logs. These logs might hold personal data such as IDs, IPs, and even names.
Reviewing your logs will allow you to find any personal data and deal with it accordingly.
Cloud Compliance
At The Bot Forge we use the Dialogflow natural language processing engine to create our chatbots. Using Google Cloud services means we can rely on GDPR being upheld with regards to our chatbot data:
At Google Cloud, we champion initiatives that prioritize and improve the security and privacy of user data. We’ve made multiple updates to ensure that Google Cloud customers can confidently use our services now that the GDPR is in effect.
We have peace of mind as compliance with the GDPR is a top priority for Google Cloud. It's important to have this confidence when using third-party services which handle your data.
Want to talk about GDPR and data privacy?
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Dialogflow Knowledge Connectors
Tips. Insight. Offers. Are You In?
In this post, I'm going to look at the new Knowledge Connectors feature in Google Dialogflow. As I look at the features in more detail I'm assuming you understand the more common Dialogflow terms and features - agents, intents & entities.
It's also important to remember this feature is in beta.
The Problem
We've been working on chatbot projects for many years now and a large number of our chatbot projects have shared a similar requirement: the ability to answer a large number of questions on a particular subject. This may be to answer technical questions about a product offering or to offer information for a particular service.
Often the information related to these types of questions is held on our chatbot customer's own websites as FAQ pages or in specific PDFs or unstructured text documents.
These types of knowledge bases can often hold large amounts of information and so technically they will provide answers to thousands of chatbot questions.
The challenge for a successful chatbot is utilising this often unstructured information to understand a question and provide the correct answer. To meet this challenge we can look at 2 approaches; the traditional one and using the new Dialogflow Knowledge Connectors.
The Solution
Traditional Approach
Stepping back a bit it's important to briefly go over the traditional approach to creating chatbot conversational ability. There are a number of different chatbot frameworks out there such as Google Dialogflow, IBM Watson, Microsoft Bot, Rasa etc and they all largely use the same concepts. A user submits a voice or text query and this utterance will be matched to an intent and any entities extracted. The matched intent would either provide a static response or rely on some form of application layer to perform the required action to provide the response to the user.
This approach can be easy. However, things can get complex and difficult to manage if the scope of intents is very large and or/ the information is constantly being updated. If we want to support questions with knowledge base information then each question needs to be created as an intent and the correct response formulated. This can lead to problems such as:
- Problems with the Intent Classification model grow causing more incorrect classifications.
- The amount of effort required to keep adding more training data to the model to ensure that the accuracy of the Intent classification remains high. Fortunately, Dialogflow provides a training UI in the web console to help keep track of any misclassified utterances, analyzing them and adding these to the training data, however, this does take time.
- Creating and managing intents to support new information in document stores.
Knowledge Connectors
Knowledge connectors are a beta feature released in 2019 to complement the traditional intent approach. When your agent doesn't match an incoming user query to an intent then you can configure your agent to look at the knowledge base(s) for a response.
The knowledge datasource(s) can be a document(currently supported content types are text/csv, text/html, application/pdf, plain text) or a web URL which has been provided to the Dialogflow agent.
Using Knowledge Connectors
To be able to use knowledge connectors, you will need to click "Enable beta features and APIs" on your agent's settings page.
Its also worth mentioning that Knowledge connector settings are not currently included when exporting, importing, or restoring agents. I'm hoping that this is something currently being put in place by the Dialogflow team.
Knowledge connectors can be configured for your agent either through the web console or using the client library that is available in Java, node.js & python. You can also configure from the command line.
To create a knowledge base from the web console, login to Dialogflow & then go to the knowledge tab. The process is fairly straightforward and involves providing a knowledge base name then adding a document to the knowledge base.
You can read more information creating a knowledge base here
After you've done that then you just need to add an intent and return the response. It's also worth keeping in mind you can send all the usual response types and that means including rich responses which I think is pretty cool.
Trying out knowledge connectors
Ok, so its time to try out these wondrous new knowledge connectors. There are 2 different types of knowledge base document type: FAQ & Extractive Question Answering. These choices govern what type of supported content can be used. There are also a number of caveats for each content type which you can read more about this here
Based on these 2 document types I looked at a couple of common use cases which we often encounter at The Bot Forge and correlate well with the document types supported:
- Chatbot FAQ functionality using an existing FAQ webpage in a fairly structured format to provide answers from.
- Chatbot FAQ functionality using information in an unstructured format to provide answers from.
I carried out my tests using a blank Dialogflow agent with beta features enabled.
1- An FAQ Knowledge Base (Knowledge Type: FAQ)
For my knowledge base I used the UCAS Frequently asked questions webpage and used the following URL as my data source. This processes the URL which is in the correct format and creates a series of Question/Answer pairs which can be enabled or disabled in the console, pretty neat!
So giving this a spin my first test was "how do I apply" and the result was spot on,
matchConfidenceLevel: HIGH matchConfidence: 0.97326803
Whilst different variations on the same question also returned a good result.
"im not sure how to apply" matchConfidenceLevel: HIGH matchConfidence: 0.9685159 "can you tell me about how I can apply" matchConfidenceLevel: HIGH matchConfidence: 0.968346
Unfortunately, when I try something a bit less obvious. I get an incorrect result as it matches the wrong intent.
"how do I submit my application" matchConfidenceLevel: HIGH, matchConfidence: 0.9626459
In this case, it's matching the "How can I make a change to my application" intent with a high confidence but unfortunately it's the wrong intent. So the problem here is we need to fine-tune the model and re-assign the training phrase (utterance) to the intended intent. The limitation is that in the knowledge base you can't fine-tune responses. If you want more control you will need to move this faq over to its own intent.
This problem is compounded by the fact that the training feature of the console just lists each response intent as "Default Fallback Intent". It's hard to check which responses have been answered incorrectly. One way round is to look in the History area of the console and look at the Raw interaction log of each response.
One really useful feature is that you can assign a specific extracted FAQ from the knowledge document and assign to an intent. Just click on view detail in the document list -> select the question and click the "convert to intents button". At the same time, this will create a new intent and disable the current Question/Answer pair. So overall pretty impressive if you have webpage or doc of structured FAQs you can use this to power an FAQ chatbot pretty effectively with some monitoring.
2-A more unstructured FAQ Knowledge Base (Knowledge Type: Extractive Question Answering)
In this use case, I wanted to try out the ability of the knowledge connectors to return answers from more unstructured data.
Again there are caveats about what data source you can use you can read more about this here.
For my test, I used a standard drug leaflet with MIME type PDF covering Priorix, from www.medicines.org.uk. I created a new knowledge base, added a new document and made sure I selected the knowledge type as "Extractive Question Answering". Once imported the PDF is listed in the document list. My aim was to validate if Dialogflow could extract some fairly simple answers from the document. Now for some testing:
"What is Priorix" matchConfidenceLevel: HIGH matchConfidence": 0.88257504 answer : "Priorix, powder and solvent for solution for injection in a pre-filled syringe Measles, Mumps and Rubella vaccine (live)"
Unfortunately, although the response had a high confidence and match score it was actually an incorrect response. Ideally, the answer should have been:
"Priorix is a vaccine for use in children from 9 months up, adolescents and adults to protect them against illnesses caused by measles, mumps and rubella viruses."
I tried another test:
"how is priorix given" matchConfidenceLevel: HIGH, matchConfidence: 0.8826 answer: The other ingredients are: Powder: amino acids, lactose (anhydrous), mannitol, sorbitol
Again this was an incorrect response. I would have expected the correct response to be:
"How Priorix is given
Priorix is injected under the skin or into the muscle, either in the upper arm or in the outer thigh."
So unfortunately not great results in extracting answers from the PDF I used. It would be interesting to look at a selection of other types of documents and corpora.
Do Knowledge Connectors work?
Again it's important to point out this is a beta feature. There are definitely challenges and in some functional areas much more to be done with Knowledge Connects. In conclusion, It's also important to recognise that I looked at 2 different types of use cases and knowledgebase document types which provided very different results so its worth looking at each one separately.
Chatbot FAQ functionality using an existing FAQ webpage in a fairly structured format.
If you want to convert your FAQ page into a chatbot or if you have a similar structured document such as a PRFAQ for a product or service then Connectors work well.
Just supplying the URL of the FAQ page as a data source to the knowledge connectors is fantastic and provides fairly good results. However, it's worth keeping in mind there may still be match errors so the history log is invaluable in checking for them. Thankfully it's fairly easy to manage any question/answer pair which has been handled incorrectly by converting to its own intent.
Chatbot FAQ using a document in an unstructured format.
I found my test results with this use case rather disappointing. The accuracy of the extracted answers was fairly poor for my test case. Although for different document sources you may be able to get better results.
The extracted answers look more like a match based on keywords with some additional coverage but it does not appear to consider the context in which the question is asked. Also, this type of knowledge connector does not provide any full control like intents in terms of context and priority of matching training phrases etc so there is no way of fixing bad responses. A feature where you can evaluate and train responses would be a great addition to the knowledge base so hopefully, that is in the Dialogflow team pipeline.
Should I use Dialogflow Knowledge Connectors?
If you have some FAQ information in a structured format then Knowledge connectors are worth a try with some caveats.
If you have unstructured documents which you want your chatbot to use to extract answers to questions then at the moment knowledge connectors are not a magic bullet. It's a big ask, but for me, this is where the real value will lie particularly if you want to support large knowledge bases with a chatbot. Knowledge connectors are an experimental feature, so hopefully as the technology advances then they will improve.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
How To Create The Perfect Google Business Welcome Message
Why Google Business Messages?
Google Maps has 155 million monthly users and it's estimated that Google handles 5.6 billion searches per day - two trillion searches a year!
Connecting with your customers at these two touchpoints is more beneficial than ever.
Gone are the days when you needed to send customers to a website or social media profile to contact you. Google Business Messages allow you to chat to customers directly in a Google search result (also known as a SERP, or Search Engine Results Page), on Google maps, or through your Google Business Profile (formerly known as 'Google My Business', or 'GMB').
With automation and chatbots for Business Messages in their relative infancy, we've included some great example welcome messages, and tips to help you get the most from your Google's Business Messages chatbot integration.
5 Tips To Help Create The Perfect Welcome Message
Improving user onboarding experiences and conversion rates of your users has never been as crucial as it is now.
A welcome message could well be the first contact a customer will have with you.
The two are unavoidably linked!
Writing better welcome messages for your chatbot solutions can really help set the tone for your customers, and establish a great experience.
These tips will help you to write more effective welcome messages for all your chatbots!
1. Greet, Introduce & Encourage
The major aims of your Google My Business chatbot welcome message should be to greet your new user, explain your chatbot's purpose, and encourage them to take action.
Use your welcome to set expectations for your customers and guide them through the initial stages of their interactions with your company.
2. Keep it Simple
Your message should only be a maximum of a few lines.
Avoid large chunks of text - try and split your content down, naturally, into separate messages.
3. Define Your Chatbot's Persona
Always make sure your users know they are interacting with a chatbot.
Humanizing your chatbot makes the whole experience more usable and fluid - if you don't convey a persona, your users will decide on it for you.
If you don't convey a persona, your users will decide on it for you
Your welcome message is the first opportunity to introduce your chatbot's personality. The language you choose sets your assistant's tone and values - which should match those of your brand.
PS - now's a good time to give your assistant a name. Keep its name short, memorable, and easy to spell & pronounce.
4. Use Emojis (But Not Too Many)
A great way to humanize your Google Business Messages chatbot is to incorporate emojis into your conversation.
Not only are emojis a great way to add space to your text, but they can also add small personality traits to match the tone of your message - whether that be amicable or exciting!
Be sure to stay true to your chatbot's tone of voice, and think about how appropriate they are in context. Emoji are difficult to use effectively in corporate or stressful/sensitive settings, e.g. funeral directors, emergency dentists, or divorce lawyers.
Whatever you go for, go easy. Too many and you'll end up looking spammy - stick to the occasional one here and there to add a visual break, rather than something to focus on.
5. Make Your Call To Action (CTA) Clear
In general, users won't have a lot of time to explore your chatbot's capabilities.
It's wise to make your call to action as clear as possible - you simply can't afford for customers to be confused when they are presented with your chatbot's welcome message.
When designing your call to action, keep your key business and marketing goals in mind.
Making use of Google Business Message's conversation starters is a good idea to follow up with, so users can easily select your CTAs.
You can use up to 5 of these, e.g. "Choose your bike" | "View bikes" | "See latest offers"
Tips. Insight. Offers. Are You In?
5 Awesome Welcome Message Examples
Now we've looked at some tips to consider, let's dive into our 5 awesome examples.
1. Conversational Commerce
This welcome is aimed at increasing sales for a bike manufacturer:
"Hi I'm {Bot’s name}, the GoodBikes virtual assistant.
I can help you, "Choose the correct bike for your riding", "View our bikes", "View latest bike offers".
How about highlighting an offer to drive sales:
"Hello, {Customer’s first name}! I’m {Bot’s name}, and I’d be happy to help you win a 25% discount on your first purchase with {Your brand’s name}. Can I help you find a particular trainer?"
"Road | Trail | Track"
2. Announcements
Offer current data that enables self-service for users.
Change the welcome message and conversation starters to temporarily contain this information if there is a big development or event that you anticipate people would look for, like a service outage.
"Hi, I'm the Livewire virtual service agent"
" I've got one thing you need to be aware of, we currently have an outage; normal service for your area is due to be resumed at 5:00 PM"
"Is there anything else I can help you with today?
3. Contextual Welcome Messages
Contextual information can be used to personalise your greeting.
The user's name, location, entrance point, and place ID are all included in your contextual data (for location-specific entry points).
For each language and place of business that you support, you're able to design a special welcome message:
"Hi I'm {Bot’s name}, the {your brands name} virtual assistant for {location}. I can answer questions about {location}, how can I help?
Closing time | Services | Parking Info"
4. Customer Support
Provide an extra level of support for your customers and manage expectations with response times.
“Hello, {customer’s first name}! Thanks for your enquiry.
Please expect a response from our support agent within 24 hours. In the meantime, why not take a look at our product tutorial: {link to tutorial}"
5. Event Specific
Your welcome messages can be seasonal or tied to promotional events; use your welcome message to highlight this to the user:
"Welcome to {your restaurant name}, I'm {chatbot name} the virtual assistant. Don't forget its burger night tonight, 20% off all burgers! What can I help you with?
Call us | Order online | Book a table"
Conclusion
Following these tips will help you stand out from the crowd and get the attention of your ideal consumers as well as help your existing ones.
Utilizing Google's Business Messages for your brand is surprisingly easy.
All you need is the help of a Google Business Messaging API partner. Knowledgeable partners, such as The Bot Forge, will help you provide rich conversational messaging solutions that will help you increase client loyalty and engagement
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
3 Books That Will Boost Your Chatbot Knowledge
Introduction
Currently, chatbots are dominating online markets, especially in countries such as the U.S., India, Germany, Brazil, and the UK.
According to a Business Insider article on chatbot statistics, 40% of internet users worldwide prefer chatbots over virtual agents because they get answers quickly and more conveniently due to their 24-hour service.
Due to the increasing demand of consumers to have round-the-clock digital experience, experts predict that retail companies will increase their budget for chatbots to $142 billion in 2024 from just $2.88 billion in 2019.
Therefore, we can expect that chatbots will become the primary communication channel for online consumers in the foreseeable future.
As such, here are three books that can jumpstart and enrich your knowledge in the world of chatbots.
1. Designing Bots: Creating Conversational Experiences
Before starting your journey, it is important to understand and familiarise yourself with the fundamental information about chatbots.
Designing Bots: Creating Conversational Experiences by Amir Shevat can help you on this one.
The book introduces the readers to the description, process, and purpose of chatbots. Shevat emphasises that chatbots are not trivial projects. Therefore, businesses that want to utilise chatbots must learn how to effectively use them to ensure a successful implementation and a return on investment.
Our previous article, Planning the Best Chatbot, concisely breaks down the necessary steps when planning a chatbot project, including identifying the purpose, goals, and performance indicators of your project. Alongside our insights, the book provides information on how to build and design your chatbot by presenting actual cases.
With the help of this book, you can learn and refresh your memory with the basic principles and start delving into the deeper concepts about chatbots without feeling overwhelmed.
2. Algorithms to Live By: The Computer Science of Human Decisions
The second book, Algorithms to Live By: The Computer Science of Human Decisions by Brian Christian and Tom Griffiths, explores the power of precise algorithms in helping us decipher human questions, which allows us to predict human choices using computer science.
The book posits that algorithms aren’t limited to the world of computers, but that they also bridge technology and human interfacing. This is why the book is worth reading.
It highlights how mathematical philosophy and life are not so different when it comes to solving problems and making decisions. By solving human problems like mathematical problems, which includes inputting various information and seeing how they work together, we can reduce and deduce possible options and end up with the right answer– a process paralleling that of chatbots'.
Solving problems and the steps that involve them are algorithms themselves, and chatbots work similarly. After customers input their concerns, the algorithm recognizes the problem and presents the most viable solution based on a predetermined set of actions or rules.
Tips. Insight. Offers. Are You In?
3. Business of Bots: How To Grow Your Company Through Conversation
This book, Business of Bots: How To Grow Your Company Through Conversation by Mariya Yao and Adelyn Zhou, goes beyond the theories of chatbots and discusses how businesses who want to connect with their customers can utilise chatbots.
This is best for those who have gone through the chatbot basics and want to advance their knowledge.
According to the authors, chatbots powered by artificial intelligence will not only help businesses improve their customer service departments, but also boost their sales and marketing strategies.
To visualise the success of chatbots, the book also examined and featured hundreds of actionable bot strategies of leading brands in the marketplace, such as Sephora, Expedia, Victoria's Secret, Capital One, and eBay.
After learning the ropes of chatbots, it is time for you to apply that knowledge and enhance your business with chatbots.
Conclusion
In a world where consumers prioritise efficiency and speed, chatbots will become the new mode of communication with consumers.
Understanding how they work and how to best utilise this technology will definitely help businesses keep up with consumer needs. If you need assistance in initiating your chatbot services, do read about our custom chatbot development services.
Post especially contributed by Cara Ariella Erickson for thebotforge.io
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
The Bot Forge is One of the Most Reviewed UK AI Companies in 2022
The Manifest Recognizes The Bot Forge as One of the Most Reviewed AI Companies in the UK
The Bot Forge creates modern solutions to improve organizational efficiency for our partners. Our team of experts helps you design, build, launch, or support enterprise-grade chatbots, voice assistants, and conversational IVR solutions. We aid you in making the most out of AI technology to enhance your business performance.
Today, we’re thrilled to share that we’re among the leaders on The Manifest this year. According to the B2B research on The Manifest, we’re one of the most reviewed AI companies in the UK.
“We are really excited to have been chosen as one of the leading chatbot and voice assistant agencies in the UK by The Manifest."
— Adrian Thompson, Founder of The Bot Forge
We are industry experts in conversational AI, and we’ve been committed to delivering AI architecture expertise to a global client base since 2018. Over the years, we’ve been successful in producing groundbreaking solutions for many organizations worldwide. This award showcases our unyielding efforts in the past years to provide advantageous technology to our partners.
In August 2022, a marketing analytics consultancy partnered with us for the development of a proof of concept. The client needed a virtual agent built on Dialogflow CX designed to illustrate the potential use case of conversational AI in the digital customer journey. The client shared the following about our partnership:
“I appreciated the creative approach because speed and lean focus are the order of the day for POCs."
— Jonathan Lewis, CEO, Marketing Analytics Consultancy
Thank you to Jonathan for taking the time to write his honest feedback. Don’t forget to browse through The Manifest, a company-listing platform, to discover more about our work.
If you’re interested in our AI solutions, please schedule a free consultation with us today.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
How Much Does it Cost To Build a Chatbot in 2023?
Tips. Insight. Offers. Are You In?
Like most software projects the chatbot cost really depends on the scale and complexity of the project.
These requirements will govern the effort involved in building your perfect automated assistant and the ongoing effort required to keep it running smoothly.
Let’s take a look at some of the governing factors in how much your chatbot project will cost.
How much does it cost to create a chatbot to fix our [insert business problem]?
We are often asked this question by clients looking to start their first conversational AI project.
At this point, we tend to ask a specific set of questions to get an idea about the scope of the project they have in mind. You can read a bit more about your chatbot scope here.
Chatbot cost can be broken down into 2 parts.
- One off design, development, deployment of the chatbot
- Ongoing monthly maintenance, hosting and management costs
Spoiler alert: we can’t give an exact price without knowing the details of your project. However, we can give some estimates based on the type of project. Jump to the end if you'd like to go straight there.
Let’s dive in and take a look at the key drivers impacting the price for custom chatbots.
Deployment Channel Chatbot Cost
The first requirement to consider is the channel in which the bot should function.
By this, I mean where the bot will be used.
It could be as a website widget implemented on your webpage or deployed into existing messaging platforms.
Examples of these platforms include Facebook Messenger, WhatsApp, Microsoft Teams, Slack, Telegram, and Viber. Some of these examples provide more complex UI elements which can be utilised in your bot.
Other obvious channels to deploy a chatbot are voice; Google Assistant, Alexa or even Conversational IVR systems which can be connected to the same conversational engine as your text-based chatbots although voice assistants will invariably need work specific to voice channels.
For our custom chatbot integrations, we normally provide one channel with the project and then charge per extra channel, as required.
We’ve found that our clients will often want a web-based chatbot first and create and deploy different versions to other channels depending on the chatbot use case.
Extra channel development costs vary, depending on the amount of work required to create the best user experience for the platform.
Features
Depending on where the chatbot will work in there is also scope to provide other functionality such as voice capability for web chatbots or enhanced chatbot interface features specific to the deployment platform.
Again, costs depend on the amount of complexity and effort involved in building each feature.
As an example, adding voice interaction capability to a web chatbot would be £3,000+.
Human-agent handover via live chat or WhatsApp integration is also a popular feature.
Natural Language Ability Chatbot Cost
If a chatbot is required to support more complex natural language understanding (NLU- you can read more about some of the tech terms here) features and not just UI elements such as buttons then this will mean that additional effort is needed to train the bot and design and implement a more complex conversational flow.
In reality, most text-based chatbots will have some level of NLU and, of course, voice assistants are all NLU.
We utilise the best of breed NLU solutions to create our conversational experiences.
In particular, we use Google Dialogflow ES and CX for many of our projects and, as a result, we are Google Tech partners and experts in Dialogflow. For onprem or open source solutions we also use Rasa.
Conversation Skills
The complexity, scope and volume of the required conversational ability also affect cost. This relates to specifics such as:
- The number of branches in the conversation tree
- The number of questions that have to be handled by the chatbot, which can often be in the thousands
- The number of training phrases needed; this element can improve the accuracy of the chatbot, (you can read more about training data here)
- The complexity of conversational ability i.e. support for complex user enquiries, multi-turn conversations.
- Number of entities needed to support the required responses, don’t forget you can remind yourself of terminology here
Languages
Chatbots are capable of supporting different languages, as long as these are supported by the NLP engine.
It’s possible to add different language permutations to the same chatbot project. However, each language will need its own testing and all the responses for each language and potentially any responses returned by business logic may need to be altered for each language.
A conversational designer will also need to consider the nuances of each language here, so costs for each language will depend on the size and complexity of the conversational ability for the primary language.
Integration Chatbot Costs
Connect your chatbot to existing systems: APIs, RPA, Knowledge Bases
The other area which will impact cost is dependent on the planned role of the chatbot: what the chatbot will need to do to carry out its role?
Will the chatbot need to integrate with current systems to provide its responses?
Will it need to hand over to live agents?
Will it need to connect with CRM and ticketing solutions?
Some chatbots may need to carry out complex interactions to provide answers to customer queries.
You can read more about possible integrations here, but the rule of thumb is that if the system you want to integrate with has an API and a means of authenticating then we can integrate with it.
Chatbots can also leverage other AI systems to provide relevant information to govern conversational flow. For example, sentiment analysis.
With so many possibilities for chatbot features, it's hard to estimate the price here as each integration can have its own complexities and sets of APIs to authenticate with.
Integration Chatbot Costs
Connect your chatbot to existing systems: APIs, RPA, Knowledge Bases
The other area which will impact cost is dependent on the planned role of the chatbot: what the chatbot will need to do to carry out its role?
Will the chatbot need to integrate with current systems to provide its responses?
Will it need to hand over to live agents?
Will it need to connect with CRM and ticketing solutions?
Some chatbots may need to carry out complex interactions to provide answers to customer queries.
You can read more about possible integrations here, but the rule of thumb is that if the system you want to integrate with has an API and a means of authenticating then we can integrate with it.
Chatbots can also leverage other AI systems to provide relevant information to govern conversational flow. For example, sentiment analysis.
With so many possibilities for chatbot features, it's hard to estimate the price here as each integration can have its own complexities and sets of APIs to authenticate with.
Deployment & Infrastructure Security
Often security demands for a chatbot project need specific features, for example, HIPAA compliance.
In these cases, SSO, RBAC, and on-prem or private cloud deployment can be used to ensure compliance with company security policies.
These can have an impact on overall project costs and again, costs are based on the demands of a specific project.
Chatbot Training & Maintenance
We offer our chatbot solutions based on a SAAS model.
Costs incurred tend to be based on a yearly subscription and again depend a lot on the scale and complexity of the chatbot.
These monthly costs will cover the following:
- Access to our world-class chatbot and voice assistant analytics platform (Chatseer)
- Daily supervised learning and improvements
- Monitoring conversations and confirming qualified intents as well as checking for unmatched intents and fixing them as needed
- Third-party and integration maintenance. Making sure your bot is performing well and healthy!
- Natural Language Understanding service costs (depending on volume and platform used).
- Hosting and data storage
- Chatbot reporting interface
- Post-development support.
As a rule, monthly maintenance costs tend to be in the region of 10% of the initial implementation cost.
Chatbot Packages
The cost of a chatbot project can vary widely depending on the overall scale of the project and the features required.
We tend to split our projects into 3 packages. You can see the features included and the one-off project costs and monthly costs in the table below.
It’s worth keeping in mind that the cost of a capable chatbot does not have to be prohibitive and it’s often easy to start small and add features as business needs require them.
We make sure all our chatbot and voice assistant projects will scale. So even if you want to start with a smaller scale chatbot solution, your company can still expand and build on this to create a large scale solution further down the line.
Discovery and requirements phase
We provide a discovery and requirements service for conversational AI projects. Our aim is to help businesses understand the potential of conversational AI and identify the requirements for a specific project.
This service typically includes an analysis of the business needs and goals, an assessment of the available data and resources, and the identification of potential challenges and opportunities.
We will work closely with you to understand the problem that needs to be solved, and to identify the key requirements for the conversational AI project. We will also recommend different conversational AI platforms and supporting technologies that can be used to build the solution.
Our goal is to help your organisation understand the potential of conversational AI and to provide you with a clear plan for how to proceed with the project. The cost of the service will depend on the complexity of the project and the level of support required from the service provider.
Deliverables
Project Cost: £3,000 to £15,000
Proof of concept (POC)
If you have an idea or use case for a conversational AI product or feature it's normally good practice to create a POC.
A POC is an early model that does not have all the final product's functionality, the main goal of a POC is to test the technical feasibility of a solution, and to identify any potential challenges or issues that would need to be addressed before moving forward and investing time and money on the development of a full-fledged system or application.
We aim to keep costs down but ensure all the work can be used as a basis for a production project.
Deliverables
- Define the problem or use case
- Identify the data and resources required
- Analyse and select conversational AI platforms and supporting technologies
- Build a simple conversational AI model
- Create integration for the chosen channel
- Carry out all required integration works
- Test the POC
- Evaluate the results
- Iterate and improve the model based on the results
Project Cost: £2,500 to £15,000
Small Project
A smaller chatbot project is a relatively simple and straightforward implementation of a conversational AI system, typically designed to address a specific use case or business need. This type of project may involve building a chatbot that can handle simple customer inquiries, such as answering FAQs or providing information about products or services.
The chatbot can be built using pre-built chatbot platform or framework, with pre-trained models that can handle natural language processing (NLP) and understand the user's intent. The chatbot can be integrated with the business's website or mobile app, and can be accessed by customers through a chat interface.
The chatbot's functionality can be limited to the specific use case that it is designed to address, such as providing customer support or information about a particular product or service. The chatbot's responses can be pre-defined, and the chatbot can be trained to understand and respond to a limited set of customer queries.
The cost of a smaller chatbot project will depend on the complexity of the use case, but generally, it is relatively low compared to a larger, more complex conversational AI project. Additionally, a smaller chatbot project will typically require less time and resources to develop and launch, and may be used as a stepping stone to larger, more complex projects in the future.
Project Cost: £2,500 to £10,000
Monthly Maintenance Cost: £200 to £1000 (depending on volumes)
- Website chatbot
- Facebook Messenger chatbot
- NLU
- 10-20 intents
- 100s training phrases
- 5 – 10 rich UI elements
- knowledge base support
- Small talk
- 1 language
Medium Project
A medium chatbot project is a more complex implementation of a conversational AI system that addresses multiple use cases or business needs. This type of project typically involves building a chatbot that can handle a wider range of customer inquiries and provide a more personalized experience.
The chatbot can be built using pre-built chatbot platform or framework, with pre-trained models that can handle natural language understanding and complex conversations. The chatbot can be integrated with multiple channels such as website, mobile app, social media platforms etc.
The chatbot's functionality can include multiple use cases, such as providing customer support, answering frequently asked questions, providing information about products or services, and even handling transactions. The chatbot's responses are generated through a combination of predefined and dynamic responses, and the chatbot can be trained to understand and respond to a wide range of customer queries.
The chatbot can also be integrated with other systems such as CRM, ERP, or inventory management systems to retrieve information and perform actions. The chatbot can also be equipped with features such as personalization, sentiment analysis, and analytics to provide a more engaging and personalized experience for the customers.
The cost and time required for a medium chatbot project will be higher than a smaller chatbot project as it involves more complexity, advanced features and integrations. However, this type of project can bring significant benefits to the business, such as cost savings, improved customer engagement, and increased efficiency.
Project Cost: £10,000 to £25,000
Monthly Cost: £500 to £2,500 (depending on volumes)
- Website chatbot
- Facebook Messenger chatbot
- Microsoft Teams Chatbot
- Slack Chatbot
- Alexa Skill
- Google Assistant
- NLU
- 1000s of training phrases
- 50-100 intents
- 10-25 rich UI elements
- Simple integration
- 1 or 2 languages
Large Project
A large scale or enterprise chatbot project is a comprehensive implementation of a conversational AI system that addresses multiple business needs and use cases across an organization. This type of project typically involves building a chatbot that can handle a wide range of customer inquiries and provide a more personalized experience, and it can integrate with multiple internal systems and processes.
The chatbot can be built using pre-built chatbot platform or framework, include extensive bespoke coding and complex integrations. The chatbot can be integrated with multiple channels such as website, mobile app, social media platforms, SMS, smart Interactive voice response (IVR), and even voice assistants like Alexa, Google Home or even digital humans.
At this level functionality can include multiple use cases, such as providing customer support, answering frequently asked questions, providing information about products or services, handling transactions, and even automating internal and external processes such as smart IVR integrations.
The chatbot can also be integrated with other internal systems such as CRM, ERP, inventory management, and HR systems to retrieve information and perform actions, as well as with external systems like payment gateways, and logistics providers.
Large scale or enterprise chatbot projects are complex and require significant resources and expertise. They can be costly, but they bring significant benefits to the business, such as cost savings, improved customer engagement, increased efficiency, and automation of internal processes. Additionally, a well-designed and executed enterprise chatbot can provide a competitive advantage and can help in differentiating the business from its competitors.
Project Cost: £25,000 to £100,000+
Monthly Cost: £2,500 to £5,000+ (depending on volumes)
- Large scale enterprise chatbot with multiple API integrations
- 10,000s of training phrases
- Personalised UX
- 1000s intents
- Multiple knowledge bases
- Deployed to multiple channels including conversational IVR
- Custom user interface elements
- Bespoke functionality
- Multiple languages
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Arabic NLP Guide [2023 Update]
Introduction
Arabic is the fourth most spoken language on the internet and arguably one of the most difficult languages to create automated conversational experiences for, such as chatbots. An Arabic chatbot is a program that can understand and respond in Arabic.
Natural language technologies enabling us to simulate and process human conversations in Arabic have improved a lot over recent years. Enabling us to train to understand the emotions, and meanings, and detect the misspellings and sentiments of the language.
In this post, we wanted to take a look at the challenges, and available tools and create a brief proof-of-concept chatbot using one of these tools.
Arabic NLP Challenges
Arabic natural language processing (NLP) is a rapidly growing field, but it also presents a number of unique challenges compared to other languages.
- Sparsity of Data: One of the biggest challenges facing Arabic NLP is the lack of large-scale, labeled datasets. This makes it difficult to train accurate models and leads to low performance on certain tasks.
- Complex Script: Arabic script is complex and includes many diacritics and ligatures, which can make text pre-processing and feature extraction more difficult.
- Morphological Complexity: Arabic has a complex morphological structure, which can make it difficult to accurately segment words and identify the root of a word. This can make tasks such as stemming and lemmatization more challenging.
- Language Variation: Arabic is spoken in many countries and dialects, which can lead to variations in vocabulary, grammar, and syntax. This can make it difficult to design models that are able to handle the diversity of the language.
- Annotation Challenges: Annotating text for NLP tasks is always a challenge, but it is even more so for Arabic due to the complexity of the language and the lack of resources.
- Right-to-Left Script: Arabic script is written from right to left, which can make it challenging to integrate with left-to-right script systems and can also affect text alignment and layout.
- Lack of Standardization: There are few standard resources for Arabic NLP, such as corpora, part-of-speech tag sets, and named entity recognition tags, which can make it difficult to compare results across different studies and to replicate previous work.
- Cultural and Religious Sensitivity: Arabic text may contain sensitive cultural and religious topics, which may require special consideration when processing and analyzing the data.
Despite these challenges, there is a lot of ongoing research and development in the field of Arabic NLP, and many organizations and researchers are working to overcome these obstacles. With the increasing demand for Arabic NLP in areas such as customer service, e-commerce, and social media, it is important to continue to invest in this field and develop solutions that can help organizations to better understand and engage with Arabic-speaking customers.
To conclude, Arabic NLP is challenging due to the complexity of Arabic script and grammar, the lack of data, and the diversity of the language.
Arabic Conversational AI Technologies
The NLP technologies include advanced machine learning algorithms, natural language understanding models, and language-specific libraries and tools which need to carry out the following tasks:
- Arabic Speech Recognition: This technology is used to convert spoken Arabic into text, which is then processed by the conversational AI system.
- Arabic Text-to-Speech: This technology is used to convert text-based input into spoken Arabic, allowing the chatbot or voice assistant to speak in the language.
- Arabic Natural Language Processing (NLP): This technology is used to understand and interpret the meaning of text written in Arabic. It includes techniques like tokenization, part-of-speech tagging, and sentiment analysis.
- Arabic Language Modeling: This technology is used to train machine learning models on large amounts of Arabic text, allowing them to understand and generate the language.
- Arabic Sentiment Analysis: This technology is used to determine the emotions and opinions expressed in Arabic text, which is useful for understanding customer feedback or gauging the effectiveness of marketing campaigns.
Technical Solutions
CAMeL Tools
CAMeL Tools is a suite of Arabic natural language processing tools developed by the CAMeL Lab at New York University Abu Dhabi.
The camel-tools package comes with a nifty ‘morphological analyzer’ which — in a nutshell — compares any word you give it to a morphological database (it comes with one built-in) and outputs a complete analysis of the possible forms and meanings of the word,
The tool will reduce orthographic ambiguity to account for several common spelling inconsistencies across dialects. Camel-tools accomplishes this by removing specific symbols from specific letters.
Repustate
The Repustate platform provides a number of natural language processing tools for analyzing Arabic dialects. It understands three major Arabic dialects – Gulf Peninsular, Egyptian, and Levantine Arabic also it Obtains granular Arabic emotion analysis by aspect rather than Visualize all the insights in a customer insights dashboard
Arabic natural language processing (Arabic NLP) powers the sentiment model, such that it differentiates between Arabic dialects while picking up on colloquialisms, language nuances, social media short forms, and even emojis.
Repustate enables you to quickly and accurately capture customer and employee sentiments to increase efficiency and improve customer experience, provides native language analysis for 23 languages, and makes social media listening effortless by seamlessly integrating with the world's most popular social networks, review sites, and news sources.
Watson NLU
IBM Watson is one of the most well-known conversational AI platforms.
IBM Watson Natural Language Understanding gives you access to detailed developer resources that help you get started fast, including documentation and SDKs on GitHub.
The Arabic Natural Language Understanding enables users to extract meaning and metadata from unstructured text data. Text analytics can be used to extract categories, classifications, entities, keywords, sentiment, emotion, relationships, and syntax from your data.
Some high-level features of the platform
- Train Watson to understand the language of your business and extract customized insights with Watson Knowledge Studio.
- Surface real-time actionable insights to provide your employees with the tools they need to pull meta-data and patterns from massive troves of data.
- Deploy Watson Natural Language Understanding behind your firewall or on any cloud.
There are some Arabic language limitations, some features are not supported in Arabic such as classifications, concepts, emotions, and semantic roles for these features.
Azure Cognitive Service
Azure Cognitive Service for Language is a new cloud-based service that provides NLP features for understanding and analyzing text.
This language service unifies Text Analytics, QnA Maker, and LUIS and provides several new features.
Most importantly it supports 96 languages including Arabic.
You can create an FAQ bot trained on unstructured data or use this to create advanced conversational experiences with the Microsoft Bot Framework.
Other Options
This is not an exhaustive list. There are many other Arabic NLP options out there (e.g Farasa, MADAMIRA, and Stanford (CoreNLP)
Botpress
Botpress is a favourite of ours as it's an all-in-one conversational AI platform.
Most importantly for this post is that the Botpress natural language understanding engine also provides Arabic natural language understanding out of the box.
Botpress is a platform that makes it easier for developers to create chatbots.
The platform assembles all of the boilerplate code and infrastructure you'll need to get a chatbot up and running, as well as providing a complete dev-friendly platform with all of the tools you'll need.
The platform contains the following features:
- To build multi-turn conversations and workflows, there's a visual Conversation Studio.
- To simulate chats and debug your chatbot, you'll need an emulator and a debugger.
- Natural Language Processing activities are built-in, including intent categorization, spell checking, entity extraction, and more.
To expand the functionality, there is an SDK and a Code Editor.
Botpress is multi-channel so your Arabic chatbot can be deployed to Slack, Telegram, Microsoft Teams, Facebook Messenger, and an embeddable online chat are among the major messaging services supported.
The platform also provides Analytics, human handoff, and other post-deployment technologies.
Botpress facilitates the creation of FAQ-style chatbots. Typically, this chatbot will rely primarily on pre-populated responses.
The platform also enables you to create more complex multi-turn conversational experiences capable of comprehending Arabic and communicating in a human-like manner. They may extract information like dates, amounts, and locations from talks.
Botpress, like any other adaptable chatbot builder platform, offers limitless bot development possibilities. Botpress may be used for almost anything, from virtual enterprise assistants to consumer-facing bots that live on popular messaging networks.
Botpress Interface Features
Although it's beyond the scope of this document to review the Botpress platform in too much detail it's useful to briefly cover the basics.
The first thing that should be mentioned is that the interface of the platform is very smooth and easy to learn in a short time, building a chatbot using Botpress is quite simple, Let's review the interfaces of Botpress.
Studio Interface
When you choose a bot, you'll be taken to the Conversation Studio. For a new chatbot, Conversation Studio creates a new flow. Update the conversational flow and train an NLU model after testing, and then test and debug the chatbot Flows
Using a user-friendly design, the Flows page assists you in creating a conversational flow.
Natural Language Understanding
Botpress is an intent-based platform. You can create intents and train the model with utterances and specify how the bot should respond. The platform also offers many of the standard NLP features:
- Entity extraction. Every phrase contains entities that help your bot understand a user’s intent and respond appropriately.
- System and custom entities. System entities are known entities that you can incorporate into your bot to accelerate development. You can also provide custom entities in the form of patterns or lists.
- Slots. These are the parameters that must be fulfilled to complete an action associated with intent. You define your slots and the NLU tags certain words from a user input that can be identified as intent slots.
- Slot filling. The engine gathers info required to satisfy a particular intent.
Q&A
The user can post frequently asked questions and their answers using the Q&A page.
Libraries
You can use hooks and actions on the Libraries page to import your custom code.
Analytics
The Analytics page shows dashboards that contain analytics data obtained during user chats.
Bot Improvement
The Bot Improvement tab helps you to monitor and develop your chatbot by managing negative comments from users.
Other Features
- Broadcast: You can use the Broadcast page to deliver information to a big group of individuals.
- Code Editor: Without leaving the Botpress Conversation Studio, you may create and update actions, hooks, libraries, configurations, and module configurations on the Code Editor page.
- HITL Next: The HITL page allows you to integrate humans into the loop of the conversation when human intervention is needed.
- Misunderstood: The Misunderstood page includes the user's input that triggered the error-handling cycle, as well as when they give negative feedback regarding the Q&A.
- Testing: You can build conversation scenarios on the Testings tab to confirm that the bot maintains its good behaviour regardless of the scenario. Unit tests are what they're called.
Arabic Chatbot POC
The intention is to build an Arabic Chatbot by using the Botpress platform which supports the Arabic language.
Botpress was chosen for this project because the easy-to-use interface and out-of-the-box functionality allowed us to create a working chatbot fairly quickly.
For this project, it's going to be an Information Provider only for a Hotel chatbot concierge. A simple FAQ Bot which is the customer will ask and the bot will respond. We used the Q&A feature in Botpress to train the bot in Arabic to understand and respond to questions.
The challenge that was faced in the early stages was that there is not enough information about the Arabic language that may help to build the best Chatbot. There is scope for more information.
Tips. Insight. Offers. Are You In?
Conclusion
There are a number of excellent natural language tools and conversational AI platforms available to create chatbots that can converse in Arabic, with the accuracy and technology of Arabic natural language understanding improving day by day.
However, there are still challenges in creating and maintaining Arabic chatbots. This is compounded by a skills shortage of Arabic speakers in the AI world who have experience in creating chatbots in multiple languages and dialects and designing conversations in these languages whilst taking each nuance of a specific language into account.
Natural Language Processing (NLP) is a challenging field and it feels like some of the major players in this space need to step up their game. Google Dialogflow and Amazon Lex are conspicuous in their absence of Arabic support.
Of course, even if Arabic NLU's strength has increased significantly, it is always possible to improve it. The NLU engines are improving all the time, and further breakthroughs are undoubtedly on the way. There will always be work to do until NLU reaches anywhere near human levels.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
9 Questions To Help Define Your Chatbot Project Scope [2023 Update]
Defining Your Chatbot Project Scope
So what is a scope of work?
In terms of conversational AI development, a scope of work typically outlines the specific tasks and objectives that will be accomplished during the project. This can include details such as the functionality of the chatbot, the technologies and platforms to be used, the process for testing and deployment, and any ongoing maintenance and support that will be provided.
So it is essentially a detailed description of the project's deliverables and serves as a blueprint for the development team to follow. The scope of work is an agreement between the client and the development agency that outlines what will be delivered and the expectations of both parties in terms of cost, timelines and quality.
The success of a chatbot or voice assistant project relies heavily on the initial planning and discovery stages. It is crucial to establish strong foundations during these stages in order to ensure a smooth progress throughout the project
Chatbot and voice assistant projects can vary greatly in terms of scope and complexity. At The Bot Forge, we often receive inquiries from customers with a wide range of conversational AI project ideas. Some have already done extensive research and have a clear set of high-level requirements in the form of use cases or user stories. Others, however, may only have a general understanding of their problem and want to explore the potential of a chatbot or voice assistant as a solution
By clearly defining the scope of the project, it will be easier to identify the resources, timelines, and budget required to successfully deliver the project
Asking the right 9 questions
Scoping a chatbot or voice assistant project is largely about asking the right questions. We often find the questions we need to ask to pin down the high-level requirements and get an idea of scope are often the same. So we try to get answers to the following questions as soon as possible:
Tell us about your business/department/area...
e.g. We are the market leaders in creating x. My team handles customer service for our organization. We have x staff members and handle approximately y queries a day. Our key metrics are z and z.
What are the business problems that you'd like to address?
e.g. “we have a problem in our company, with getting too many questions about x via y, and they are often the same. What can a chatbot do to help us? Where can it be used?"
What are your objectives?
e.g. we want to use a chatbot to answer these repetitive questions so our team can concentrate on answering the more difficult queries
What is the expected process flow and/or user journey of the chatbot?
e.g. Our users request information via phone call/email/IM. 80% of these questions are simple and repetitive. We want to address these key queries via a chatbot. Further, we wish to provide our users with a simple way to speak to a live agent if more information is needed.
What are the use cases that the chatbot or voice assistant needs to satisfy and how complex are they?
e.g. we have a handful of questions which we get asked a lot. These are standard customer support requests such as where is my order? These can be answered in a few minutes. We also receive 50 or 60 other inquiries about our products which can be very detailed.
What channels does the chatbot need to be deployed to?
e.g. the chatbot will need to be deployed onto our department website as a web widget and a Microsoft Teams chatbot.
Do you need to support hand over to live agents to answer advanced customer queries or to step in when needed?
e.g. there are live agents which we would like to hand over the chat conversation to
What are the additional requirements and/or conditions?
e.g. we need to log a ticket detailing the chat into our system. We need a Spanish language version
How are you going to measure success?
e.g faster response to the customer, reduced hours spent answering questions, NPS
Once we have the answers to these questions, businesses invariably want a rough estimate: how long will the solution take to build? How much will that cost? We find we are now in a better place to give some high-level estimates. We are also in a position where we understand what our customer's problems are and how we can address them with conversational AI.
TIP: Download our free chatbot checklist to help you iron out your scope
Digging into more detail for chatbot project scope
Building on our standard questions it's then time to look in more detail at your scope:
- Purpose: Clearly define the purpose of the chatbot, such as automating customer service, providing information, or completing transactions.
- Functionality: Identify the specific tasks that the chatbot will be able to perform, such as answering frequently asked questions, booking appointments, or processing orders.
- User flows: Define the user flow and interactions that the chatbot will have with the user, including the type of input and output (text, voice, etc.).
- Integrations: Identify any systems or APIs that the chatbot will need to integrate with, such as a CRM or inventory management system.
- Languages and tone: Define the language and tone that the chatbot will use to communicate with users.
- Data and security: Define the data that the chatbot will access and the security measures that will be put in place to protect that data.
- Channel Deployment: Identify the platforms and channels where the chatbot will be deployed, such as a website, mobile app, or messaging platform.
- Ongoing maintenance and support: Identify the ongoing maintenance and support that will be required to keep the chatbot running smoothly, such as bug fixes and updates.
- Performance metrics: Define the performance metrics that will be used to measure the chatbot's success, such as response times and customer satisfaction.
- Timeframe: Define the project's start and end dates and the milestones that will be reached along the way.
Tips. Insight. Offers. Are You In?
You Have Your Scope... What Next?
Once you have your chatbot scope it’s time to work closely with your chatbot solutions provider (*tip*, that’s The Bot Forge!) in a more detailed project planning phase which can be broken down into the following steps:
- Sure, here are the steps for the project planning phase of a conversational AI project such as a chatbot or voice assistant:
- Define project goals, and objectives: Clearly identify the project's purpose, what it will deliver and what it will not deliver. Define specific, measurable, achievable, relevant and time-bound (SMART) goals to align project objectives with the overall business objectives.
- Identify stakeholders and establish communication plan: Determine who will be impacted by the project, who will be involved in the project, and who will be responsible for approving and implementing the project. Create a communication plan to keep stakeholders informed of project progress, decisions, and issues.
- Assemble project team and assign roles and responsibilities: Identify the resources required to complete the project, including internal team members and external vendors or contractors. Assign specific roles and responsibilities to team members to ensure clear accountability and effective collaboration.
- Create a project plan and schedule: Develop a detailed project plan that outlines the project's activities, dependencies, milestones, and deliverables. Create a schedule that shows the start and end dates for each activity, and allocate resources accordingly.
- Identify risks and develop a risk management plan: Identify potential risks that could impact the project's success, such as budget constraints, schedule delays, or resource shortages. Develop a risk management plan to mitigate or avoid these risks.
- Obtain approval and funding: Once the project plan is complete, present it to stakeholders for review and approval. Obtain the necessary funding to support the project.
- Implement the project plan: With the project plan approved, project team can start to implement the plan and execute the project activities as per the schedule.
- Monitor and control the project: Continuously monitor the project's progress against the plan, identify and resolve issues, and make adjustments as needed.
- Close the project: Once the project is completed, document the results, and perform a post-project review to identify what worked well, what could be improved, and what lessons were learned.
Requirements analysts, Product managers, and stakeholders, conversation designers, AI trainers and developers all need to be involved in the planning stages. The key to a successful project is to ensure the whole team has a shared vision of what the problem is, what the solution is, and who the solution is for. It's this mutual understanding which will ensure the successful creation of a high-level plan and the bedrock for a successful conversational AI project.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
What’s The Difference Between ChatGPT & GPT3?
ChatGPT Confusion
At the time of writing, since the launch of ChatGPT at the end of November 2022, numerous solutions have hit the market claiming to be ChatGPT-branded. However, I'm here to clarify that these solutions are not ChatGPT, but rather GPT3 solutions.
There seems to be a lot of confusion between ChatGPT and GPT3.
This is compounded by some solutions in the marketplace claiming to use ChatGPT which is either totally wrong, trying to ride the hype train, or genuinely confused with the underlying tech.
So, What's The Difference Between ChatGPT & GPT3?
If you got this far, you're probably wondering, what the differences are between ChatGPT and GPT3.
GPT3 stands for Generative Pre-trained Transformer, which is a Large Language Model (LLM) built by OpenAI and released in June 2020. The GPT3 model was later iterated into GPT3.5, also known as InstructGPT, to improve its ability to follow instructions and complete tasks.
If you are using the Davinci model from OpenAI's API, you are using InstructGPT or GPT3.5.
Many companies have already used GPT3 and GPT3.5 to enhance their existing products and create new ones, such as AI-assisted writing tools. However, these are not ChatGPT.
ChatGPT has undergone further training, including RLHF (Reinforcement Learning from Human Feedback).
The training process involved humans reviewing and rewriting responses generated by the model to make them more factually accurate and conversational.
The responses were then fed back into the model to train it on how to produce more human-like answers.
The model then underwent a reward model training process where multiple responses were generated and ranked by humans based on their quality and fit.
The data was then fed back into the model to train it on what constitutes a good response.
OpenAI also used Proximal Policy Optimization (PPO), a Reinforcement Learning approach, to create policies for the ChatGPT language model.
This process not only improved the accuracy and conversational breadth of responses but also trained the model to produce safer content by blocking racist, sexist, or inappropriate responses.
However, despite being built on the GPT3.5 model, ChatGPT produces better responses than GPT3.5, as it has been specifically trained to generate high-quality, human-like responses.
This is why people are often confused between the two models.
What Can You Do With GPT3?
GPT3 is incredibly powerful and ultimately the foundation of ChatGPT. Its applications are wide-ranging because it has an API to build plugins or tools that use Large Language Models.
It can generate text, answer questions, translate languages, summarize long articles, and even complete tasks such as coding and creating charts.
Here are a few things you can do with GPT-3:
- Text generation: GPT-3 can be used to generate human-like text in various styles, from creative writing to news articles, and even poetry
- Question answering: GPT-3 can answer a wide range of questions with high accuracy, making it a useful tool for knowledge management and customer service
- Chatbots: GPT-3 can be used to develop advanced chatbots that can handle complex conversational tasks, such as booking a flight or ordering food
- Language translation: GPT-3 can translate text from one language to another, providing near-human-level accuracy
- Content summarization: GPT-3 can summarize long articles or documents into concise summaries, making it easier to quickly understand the most important information
- Code generation: GPT-3 can write code, from simple scripts to complete applications, making it a valuable tool for software development
- Creative applications: GPT-3 can be used for creative projects, such as generating music, visual arts, or even video game design.
What Can You Do With ChatGPT?
As I've covered previously ChatGPT is built to have conversations and do it well. Its conversational capabilities (remember only its own chat interface) are better than GPT3 because it's been tuned to do it. It understands natural language input and generates human-like responses to questions and supports follow-up questions.
Some of the things that can be done with ChatGPT include:
- Question answering: ChatGPT can answer a wide range of questions, providing users with relevant and accurate information on a variety of topics
- Conversation simulation: ChatGPT can mimic human conversations, making it ideal for use in customer service, virtual assistance, and other scenarios where a human-like interaction is required.
- Text generation: ChatGPT can be used to generate text, such as product descriptions, headlines, and other types of content
- Summarization: ChatGPT can be used to summarize long articles, news reports, or other written content into shorter, more concise text
- Translation: ChatGPT can be used to translate text from one language to another, making it a useful tool for global communication.
Summary
So.... ChatGPT is built on the GPT3.5 model but is a conversational interface accessible only through a browser and does not have a publicly available API..at the time of writing. One thing we know about conversational AI is that the landscape is changing quickly and an API release is coming soon.
GPT3.5 is available via an API and a browser and it's this technology that can be used to create lots of different applications including chatbots.
It's also worth mentioning that GPT3.5 can also be leveraged to provide conversational experiences "Like" ChatGPT. Remember they are not ChatGPT, but it's possible to leverage GPT3.5 and add enhanced conversational capabilities which are similar. So memory and context coupled with a smaller domain area knowledge, not the entire internet!
This type of implementation is possible and is something we are working on at The Bot Forge so feel free to get in touch if you'd like to learn more.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
What is ChatGPT?
AI-Mazing
ChatGPT is the latest technology release from the team at OpenAI and it's taken the internet by storm. Reactions ranged from amazement to scepticism, with everything in between.
It's been hugely popular with more than a million people using it in the first 5 days.
ChatGPT is OpenAI’s latest large language model, released on Nov 30th 2022, in beta as a chatbot app which you can interact with on the OpenAI website.
Like GPT3, it can explain scientific and technical concepts in a desired style or programming language, brainstorm basically anything you can think of asking... and yes, of course, hold pretty complex conversations!
Like any new technology in this field, particularly one which has caused such a stir, it's important to not get too over excited and consider the real value, what can it do and what are its limitations.
Let's take a look...
So, What is ChatGPT?
You're probably wondering how it works.
The system is built on a type of artificial intelligence (AI) called a large language model. The ChatGPT model is based on reinforcement learning from human feedback and the GPT series of models from OpenAI which are themselves trained on extraordinarily large amounts of data.
To create ChatGPT, the latest GPT-3.5 Instruct model was fine-tuned with conversation examples, instead of the whole internet, to concentrate on improving the model's specialist conversation abilities. Reinforcement learning was then used so that the model could practice its conversation skills and improve.
The system works like any online chatbot, and you can simply type out and submit any question or prompt you'd like the AI to address.
The ChatGPT service is currently accessible via a web chat interface. Its simple interface hints at what it can do but also some of its limitations.
What Can ChatGPT Do?
ChatGPT can do a hell of a lot. It's beyond the scope of this post, or any post for that matter, to cover everything it can do, but let's focus on some of the things it does well.
It's also worth mentioning that all of these tasks are achieved automatically. No need to write complex prompts, configure temperature, or fine-tune models.
It can have complex conversations with you
The new chat capability of ChatGPT allows for a pretty natural conversation experience.
It is the first chat-focused large language model and it's really good at it.
Previous attempts at creating GPT3-based chat experiences has been hacky and pretty unreliable.
It has long-term memory of up to 8192 tokens and can take input and generate output about twice as long as GPT3.
All of this allows ChatGPT to maintain context and generate follow-up responses with astounding ease.
The ability which ChatGPT has to hold a multi-turn conversation is, in my opinion, the real game changer here.
To create a similar experience using the latest Intent-based conversational AI technology, i.e Rasa, involves a serious amount of work.
It can have safer conversations
In order to make ChatGPT "safe," OpenAI spent a lot of time attempting to disable responses that dealt with violence, terrorism, drugs, hate speech, dating, sentience, and eradicating humanity.
OpenAI also claimed to disable web browsing and knowledge of current dates.
It's still relatively easy to compromise these security efforts though. OpenAI have placed clear warnings that this is a test and encourages reporting of results.
So you can assume this feedback will be used to make service safer.
It's incredibly creative
As we've previously discussed, the best use cases for ChatGPT involve creative writing, brainstorming, drafting, and creative information presentation.
Any tasks where originality is valued more highly than accuracy.
Writing blog posts and content
From blog posts, to summarisation, to copywriting, to writing lecture notes about complex subjects... ChatGPT is pretty good at taking on any of these tasks.
Normally after a little bit of polishing and editing the end result is really impressive.
Writing lyrics and short stories
It's also amazing at generating rhyming poetry and producing lyrics and understanding the components of a song i.e chorus, verse, and bridge.
This is a huge leap compared to GPT3 where, from our experience, you have to finetune a DaVinci model based on a specific lyrics dataset.
It Can Help Create Better Conversational Experiences
ChatGPT and LLMs can do a lot to help create better chatbots and voice assistants.
This is really interesting for the world of conversational AI.
Whilst some people in the industry are looking nervously at their intent-based models, others are imagining a better world where LLMs such as ChatGPT can really help them to create better chatbots and voice assistants.
Here are some examples of what ChatGPT can do:
- Creating intent utterances
As a tool for writing utterances for intents, it's fantastically capable of providing training data for intent matching. In the absence of any available customer conversational data, ChatGPT utterances can serve as a good means to bootstrap a chatbot ready for further iterations.
- Entity generation
Create entities by asking for permutations e.g "list 10 different ways of saying desktop" or "list 10 different types of cycling"
- Prompt variations
It's good practice to create different prompts to make for a better experience, you all knew that right? Well ChatGPT is brilliant for creating different ways of saying the same thing: "give me 5 ways of saying would you like a tip of the day"
- Happy path creation
ChatGPT can help provide some conversation path examples if you provide it with an outline of the chatbot/VUI you are creating. We've been using this to create happy paths for Alexa skills but this approach could be used for any conversational experience.
- Persona creation
ChatGPT can also be used to simulate conversations and provide more detail about specific personas
- Test case generation
This is a really useful one. If you want to create test utterances to train a model then for each specific intent you can simply ask ChatGPT to create some for you. Just explain the intent itself and away you go
What's even more useful is the ability to prompt ChatGPT to create long tail test utterances with a simple follow-up question to the previous one.
It Can Write Code Examples
There are already some really interesting software products based on GPT3 Codex with one of the most standout apps being GitHub Copilot.
ChatGPT continues this with the ability to create pretty good code output. Here is an example of a request for some simple Javascript code. Note the fantastic formatting capabilities of ChatGPT:
ChatGPT even has the ability to create test code "can you create some unit tests in jest to test the object" and even to debug code to a limited degree.
What are ChatGPT's Limitations?
ChatGPT's drawbacks are highlighted nicely on the ChatGPT UI:
- May occasionally generate incorrect information
- May occasionally produce harmful instructions
- Limited knowledge of world and events after 2021
This is a pretty good summary.
ChatGPT Sounds like many of the millions of people posting on Twitter; they sound confident but can still be wrong.
The security systems which are obviously a large part of OpenAI's roadmap for the technology can still be fairly easily broken.
Yes, it's a large language model which has only been trained on historical data so there is a limitation there if you want to provide information.
However, ChatGPT is aware of its own limitations. If you ask it about the current situation in Ukraine:
"I'm sorry, but I do not have current information about the situation in Ukraine. My training data only goes up until 2021."
What's Next For ChatGPT?
We feel there is still a lot to come. Definitely including improved safety and we feel there are more features to come. There have already been features added including conversation history and improved performance.
The brilliant formatting capabilities mean ChatGPT could be moving towards the ultimate encyclopedia, although the risks of misinformation are still there.
The costs of a technology like this are eye-watering, so we're pretty sure that at some point ChatGPT is going to cost. There are already signs of this from the daily platform limits being introduced in the latest release.
There is no doubt that there is some big money here. At the time of writing, ChatGPT has been valued at $9 Billion, and OpenAI's valuation may be approaching $30 Billion.
It remains to be seen how much OpenAI will charge for ChatGPT when it's finally released into production.
Conversational Search
There has been a lot of discussion about conversational search products and whether they will soon rival the big search engines. Many people immediately saw ChatGPT as superior to Google search. However the reality is that they are very different. The biggest points to make are that the information from ChatGPT is often not correct, responses fabricated and not up to date.
At the moment there are a lot of possibilities for conversational search and large language models and it's likely that the big search engine providers will be focusing on this area over the course of 2023.
A note on watermarking
The quality of the output from ChatGPT has already led to students and others passing off ChatGPT as human-generated. This is being treated as seriously as plagiarism.
Because of the rise in "AIgiarism" (AI-assisted plagiarism) there are increased calls for ways to identify where AI is used to generate content.
Metadata watermarking looks like a viable option to combat this.
There are already OpenAI detector tools on Huggingface. OpenAI has a working prototype of the watermarking scheme that “seems to work pretty well", according to an OpenAI researcher.
It's suggested that a few hundred tokens – or a paragraph of text – is the point needed to get a reasonable signal that the text came from GPT3.
There is going to be more demand for this sort of technology to tackle Algiarism, mass propaganda generation, or writer impersonation.
Conclusion
It's pretty mind-blowing and I've really only scratched the surface of what's possible in this post.
ChatGPT and LLMs look set to change the conversational AI landscape forever.
If OpenAI couples the ability to fine-tune models based on ChatGPT with your own knowledge bases then this would enable the creation of a conversational FAQ with ChatGPT's engaging conversational abilities which would really open up some fantastic possibilities.
There will undoubtedly be an increase in the use of intent based and LLM conversational AI experiences.
And finally, it's really important to note that, at the time of writing, the ChatGPT service is in beta, and not production ready. It's only available for human use, via the OpenAI UI.
There is no API to use to talk to ChatGPT, unlike the other OpenAI models which are easy to use and integrate into existing workflows and tools.
For now, sit back and figure out your next question for ChatGPT.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Build a Custom AI-Powered GPT3 Chatbot For Your Business
The Rise Of ChatGPT
ChatGPT has rapidly gained popularity worldwide, with millions of users relying on its vast knowledge database and ability to hold a multi-turn conversation of considerable complexity.
However, despite its usefulness for general information, ChatGPT is limited to pre-2021 publicly available internet data, and it has no access to your private data or recent sources of information. Imagine how beneficial it would be to your business if something like ChatGPT had access to this information.
There has been considerable demand for chatbots similar to ChatGPT but without these limitations. Let's take a look at what is possible.
What is a Large Language Model (LLM)?
A large language model is a type of artificial intelligence model that is designed to understand and generate human language.
It's built using machine learning techniques and is trained on massive datasets of natural language text to learn the patterns and rules of language. These models use deep learning algorithms, such as neural networks, to understand the relationships between words, phrases, and sentences, allowing them to generate text that sounds natural and human-like.
One of the most well-known examples of a large language model is GPT-3 (Generative Pre-trained Transformer 3), which was developed by OpenAI and has become one of the most powerful language models available.
GPT-3 is particularly powerful because it has been trained on an enormous amount of data - over 570GB of text - making it one of the most sophisticated language models currently available.
Other Large Language Models Are Available
Although we've only referred to the GPT-3 large language models provided by OpenAI, it's worth noting that there are a number of other LLMs. Each LLM has its own strengths and weaknesses when asked to process and understand natural language in various ways; here are some examples:
- T5: T5 (Text-to-Text Transfer Transformer) is a large-scale language model developed by Google. It was trained on a diverse range of text-based tasks and can perform various natural language processing (NLP) tasks such as text summarization, question-answering, and translation.
- BERT: BERT (Bidirectional Encoder Representations from Transformers) is another large-scale language model developed by Google. It can be used for various NLP tasks, including question-answering, sentiment analysis, and text classification.
- RoBERTa: RoBERTa (Robustly Optimized BERT Pretraining Approach) is a large-scale language model developed by Facebook AI. It is an improvement over BERT and performs better on several NLP benchmarks.
- XLNet: XLNet is another language model developed by Google that uses an autoregressive approach for language modelling. It achieves state-of-the-art performance on several NLP benchmarks.
- GShard: GShard is a distributed large-scale language model developed by Google that achieves state-of-the-art performance on several NLP benchmarks. It is trained using a novel hierarchical approach that enables it to scale to trillions of parameters.
- Bloom: Bloom, said to be the world's largest open multilingual language model, is one of the latest LLMs and is available via the Hugging Face platform.
Access to these models is also available via other providers e.g Cohere, GoogleAI, and Hugging Face.
The takeaway here is that LLM technologies are becoming increasingly accessible to create tailored conversational experiences.
Are we using ChatGPT or GPT-3?
It's easy to get confused between ChatGPT and GPT-3, something we've looked at before in detail.
GPT stands for Generative Pre-trained Transformer, which is a Large Language Model (LLM) built by OpenAI and released in June 2020. The GPT3 model was later iterated into GPT3.5, also known as InstructGPT, to improve its ability to follow instructions and complete tasks.
What makes GPT-3 so groundbreaking is its ability to generate natural language text that is virtually indistinguishable from text written by humans. The model is trained on an incredibly large dataset of internet text, including books, articles, and websites, which allows it to understand the nuances of human language and generate responses in a natural, conversational style.
On the other hand, ChatGPT is built on top of GPT3 but has been enhanced with further training processes.
What we are going to be looking at in this post is using GPT-3 to create an experience similar to or like ChatGPT... but not using ChatGPT!
Use Your Data To Power a GPT3 Chatbot
Part of the challenge of creating a large language model chatbot based on your organisation's data is accessing the data and loading it into the correct form for it to be used in the data ingestion process.
Increasingly, businesses store their knowledge in various locations, depending on the type of knowledge and the company's specific needs and across a range of formats. However, more often than not this unstructured text data will be in a form that you can work with. Some good examples include:
- HTML
- PowerPoint
- Podcast content
- YouTube video transcripts
- Internal databases
- Customer support queries
- Other APIs
- Documentation sources e.g GitBooks
Once a data source has been identified and extracted, the next stage is to clean and preprocess the data to ensure that it is in a format that can be used. This process may involve removing duplicates, cleaning and labelling text, and standardizing formatting to ensure consistency across different data sources.
This data can then be processed and used in your interactions with your large language model.
Technologies To Interface With LLM
Whilst it's entirely possible to code up a solution to interact with LLMs from the ground up that is also time-consuming and complex. There is a growing list of offerings that can help achieve your conversational AI use-case goal.
The technologies for orchestrating chatbots based on LLMs like GPT3 are evolving rapidly. These technology stacks provide the tooling we need to create a conversational engine that can interact with LLMs easily.
Orchestration
Functionality across the different platforms falls into the same categories of existing conversational AI platforms with offerings falling into the classes of Pro-Code, Low-Code and No-Code solutions.
There are a number of these tools/platforms currently available e.g Dust, Langchain - each could warrant a dedicated post.
It's a bit of an oversimplification of what these technologies actually do but as a summary, they provide the features needed to carry out the steps needed to create conversational use-cases such as chatbots, text generation, and Q&A by interacting with a LLM.
Features
For our use case, we are looking to create a conversational agent similar to ChatGPT so the following features all come into play:
- Tools to make sense of large volumes of unstructured text data
- Tools to work with Vector stores
- Prompt generation assistance (A prompt is an input to a language model, a string of text used to generate a response from the language model).
- Accessible wrapper to talk to your LLM of choice
- Tools to enable the management of conversation state and context
Vector Stores
A vector store is a specific type of database optimized for storing documents, and embeddings, and then allowing for fetching the most relevant documents for a particular query.
These are important for our GPT3 knowledge-base powered chatbot as they store our document embeddings as indices for the search. Notable libraries are the FAIIS open-source library and the Weaviate open-source vector search engine
Creating a GPT3 Chatbot
For this example, we'll look at using Langchain to create our GPT3 chatbot. Of the platforms mentioned earlier, Langchain is our favourite. It's pro-code but is well-supported with examples and documentation.
We've included high-level technical detail in this guide; here are the main steps.
- Source your data: Take your unstructured text and clean it and prepare it for use.
- Chunk/Embed text/Load embeddings: Load your text into smaller pieces. Convert each chunk of text into a numerical format so that you can find the most relevant chunks for a given question. Put the numerical embeddings and documents into a vector store, which helps to quickly find the most similar chunks of text to a given question.
- Prompt engineering: Create the correct prompts to pass onto the LLM based on context, question history and required behaviour.
- Deploy your chatbot: Integrate the chatbot service into your channel, this could be really simple... or a slick chat UI similar to ChatGPT.
- Talk to your chatbot: Once deployed, you can start asking questions. When a user submits a question, the chatbot will identify the most relevant chunks of text in the vectorstore and generate a response based on that information and the current context and conversation state.
- Fine-tune the model: As the chatbot is used, you may find that it is not always generating the best responses. In that case, you can fine-tune the model by adding more data or adjusting the weighting of different chunks of text in the vectorstore.
- Evaluate the performance: To ensure that the chatbot is working as intended, you should regularly evaluate its performance. This may involve analyzing user feedback, monitoring the accuracy of its responses, or testing it against a range of different queries and of course keeping your vectorstore up to date with any new information.
What Is The Result?
Results
As a very quick POC, we ingested all the text from thebotforge.io and ran it through our process. The results are actually pretty good.
A GPT3 chatbot project created using the latest LLM orchestration stack provides good results on an unstructured dataset e.g website contents. It handles context, so follow-up questions about the ingested data work pretty well.
We can ask specific questions about the subject matter in a number of different ways, and ask follow-up questions.
To be honest, it's not as good a ChatGPT, but we wouldn't expect it to be. Its capabilities aren't as wide-ranging, but that is perhaps the point - they don't need to be. The main capability is that it provides the ability to talk about your knowledge base with much more flexibility than an intent-based conversational AI experience, which would take a lot longer to create and would be unlikely to be anywhere near as powerful.
Hallucinations
LangChain helps to overcome hallucinations which is an issue with LLMs.
In the context of large language models (LLMs), "hallucinations" refer to when the model generates text that is not coherent, relevant, or accurate. Hallucinations can occur when the LLM generates text that is not based on the input or task at hand but is instead based on its own learned patterns or biases.
This can happen because the LLM has learned certain patterns in the data that do not apply to the specific context of the task. In other cases, an LLM may generate text that is completely unrelated to the input or task, which can be described as "hallucinating" text.
To mitigate the risk of hallucinations, LLMs need to be trained on high-quality data, and the generated text needs to be evaluated to ensure that it is relevant and accurate to the task at hand. This can be handled within the conversation itself e.g highlighting knowledge-base content with chat responses which is where tools like Langchain come in.
Limitations
OpenAI Services
There are limitations related to the OpenAI service.
One is the cost of interacting with OpenAI's models. In the case of using OpenAI for our Langchain example, this could get expensive pretty quickly as we are using Davinci 003 which is the most capable of their current models, but also the most expensive. We also found we are running close to the maximum prompt size for our interactions.
The second issue is that we found the API calls can be laggy at times, which means poor performance for the chat interface, and more worryingly we received rate limit errors from the service because of high traffic.
Transactional Chat
No intents and integrations here. So where the ability to handle free conversation is good if a user of your GPT3 chatbot is at a stage of a conversation where they need to carry out a specific task, then this is where your chat service would need to hand over to a more intent-based approach.
We've found that a blended approach of LLM & Intent-based service works well here. Catch a support intent from a user then hand them over to your conversational AI intent service or live chat agent to manage the transaction... you can even hand them back once it's complete.
Conclusion
There is no doubt that ChatGPT has gained a huge amount of traction over a short space of time, but it's worth remembering that it's based on GPT3 technology which has been around for a while.
Despite the wonder of ChatGPT's ability to follow a line of conversation any number of times about any number of subjects, it still has obvious limitations. The most notable being there is no API (at the time of writing), it's trained on data up to 2021 and it has no real knowledge of your organisation's recent or private data.
Let's not forget that the essence of any LLM's functionality is to produce a reasonable continuation of whatever text it's got so far. It's ChatGPT's conversational "qualities" which you could argue have driven its popularity.
To handle questions about a subject or domain specific to your organisation then a GPT3 or other LLM-powered chatbot makes a lot of sense, particularly when you can give it similar "qualities" to ChatGPT.
Overall the future looks bright for LLM-powered conversational experiences. Technology in this space is progressing rapidly with a ChatGPT API in the pipeline and with rivals to ChatGPT already planned e.g Hugging Faces' next version of the BLOOM LLM.
It's also going to be the continued advancements in smaller scale fine-tunable streamlined LLMs and automated NLP model compression and optimisation tools which will begin to power a lot of our chatbot conversations.
If you want to talk to us about leveraging AI and your organisation's data, get in touch.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Build a Custom AI-Powered GPT3 Chatbot For Your Business
The Rise Of ChatGPT ChatGPT has rapidly gained popularity worldwide, with millions of users relying on its vast knowledge database and ability to hold a multi-turn conversation of considerable complexity. However, despite its usefulness for general information, ChatGPT is limited to pre-2021 publicly available internet data, and it has no access to your private data…Read More
What’s The Difference Between ChatGPT & GPT3?
ChatGPT Confusion At the time of writing, since the launch of ChatGPT at the end of November 2022, numerous solutions have hit the market claiming to be ChatGPT-branded. However, I’m here to clarify that these solutions are not ChatGPT, but rather GPT3 solutions. There seems to be a lot of confusion between ChatGPT and GPT3.…Read More
What is ChatGPT?
AI-Mazing ChatGPT is the latest technology release from the team at OpenAI and it’s taken the internet by storm. Reactions ranged from amazement to scepticism, with everything in between. It’s been hugely popular with more than a million people using it in the first 5 days. ChatGPT is OpenAI’s latest large language model, released on…Read More
Arabic NLP Guide [2023 Update]
Introduction Arabic is the fourth most spoken language on the internet and arguably one of the most difficult languages to create automated conversational experiences for, such as chatbots. An Arabic chatbot is a program that can understand and respond in Arabic. Natural language technologies enabling us to simulate and process human conversations in Arabic have improved…Read More
9 Questions To Help Define Your Chatbot Project Scope [2023 Update]
Defining Your Chatbot Project Scope So what is a scope of work? In terms of conversational AI development, a scope of work typically outlines the specific tasks and objectives that will be accomplished during the project. This can include details such as the functionality of the chatbot, the technologies and platforms to be used, the…Read More
How Much Does it Cost To Build a Chatbot in 2023?
Tips. Insight. Offers. Are You In? Your Email Address Please enter a valid email address. I agree that The Bot Forge can email me news, tips, updates & offers. I know that I can unsubscribe at any time. You must accept the Terms and Conditions. Sign Me Up Thank you for subscribing! Something went wrong.…Read More
The Bot Forge is One of the Most Reviewed UK AI Companies in 2022
The Manifest Recognizes The Bot Forge as One of the Most Reviewed AI Companies in the UK The Bot Forge creates modern solutions to improve organizational efficiency for our partners. Our team of experts helps you design, build, launch, or support enterprise-grade chatbots, voice assistants, and conversational IVR solutions. We aid you in making the…Read More
How To Create The Perfect Google Business Welcome Message
Why Google Business Messages? Google Maps has 155 million monthly users and it’s estimated that Google handles 5.6 billion searches per day – two trillion searches a year! Connecting with your customers at these two touchpoints is more beneficial than ever. Gone are the days when you needed to send customers to a website or social media profile…Read More
3 Books That Will Boost Your Chatbot Knowledge
Introduction Currently, chatbots are dominating online markets, especially in countries such as the U.S., India, Germany, Brazil, and the UK. According to a Business Insider article on chatbot statistics, 40% of internet users worldwide prefer chatbots over virtual agents because they get answers quickly and more conveniently due to their 24-hour service. Due to the…Read More
6 Tips to Ensure Your Chatbot is GDPR Compliant
Tips. Insight. Offers. Are You In? Your Email Address Please enter a valid email address. I agree that The Bot Forge can email me news, tips, updates & offers. I know that I can unsubscribe at any time. You must accept the Terms and Conditions. Sign Me Up Thank you for subscribing! Something went wrong.…Read More
What Makes a Successful Chatbot Project?
Tips. Insight. Offers. Are You In? Your Email Address Please enter a valid email address. I agree that The Bot Forge can email me news, tips, updates & offers. I know that I can unsubscribe at any time. You must accept the Terms and Conditions. Sign Me Up Thank you for subscribing! Something went wrong.…Read More
10 Questions To Ask When Planning a Chatbot Project
Tips. Insight. Offers. Are You In? Your Email Address Please enter a valid email address. I agree that The Bot Forge can email me news, tips, updates & offers. I know that I can unsubscribe at any time. You must accept the Terms and Conditions. Sign Me Up Thank you for subscribing! Something went wrong.…Read More
Building a Chatbot Using Amazon Lex
What is Amazon Lex Amazon Lex is a service by AWS for building conversational interfaces into any application using voice and text. Lex has quickly become popular among chatbots enthusiasts looking to leverage the technology which powers Alexa. Users can be up and running within minutes with no upfront costs. Amazon Lex has been in…Read More
6 Common Mistakes to Avoid When Developing a Chatbot
We are experts in developing chatbots so we know If you are looking to streamline certain operations of your business, developing a chatbot is a great way to go about it. After all, you can use technologies such as artificial intelligence (AI) and natural language processing (NLP) so that it can follow different types of…Read More
Buckinghamshire Business Festival Sponsor 2021
We are proud to be a Buckinghamshire Business Festival sponsor this year We are proud to be sponsoring the 2021 Buckinghamshire Business Festival, running from April 19th – 30th. The festival has been organised by Buckinghamshire Business First, with a packed schedule of events and opportunities to make new connections across the two weeks. Look…Read More
IBM Watson for Building Chatbots
IBM Watson
Developer tools that make it easy to incorporate conversation, language, and search into your applications. Watson gives you access to detailed developer resources that help you get started fast, including documentation and SDKs on GitHub.
There are several IBM Watson APIs available on the IBM Cloud. One of them is IBM Watson Assistant. Watson Assistant enables you to build apps that include natural language processing and structured conversation. The service provides an API which you can call from an app or website to hook into your chatbot.
Watson Assistant API can:
- Extract meaning from natural language
- Discover patterns in data sets
- Understand the "tone" of language
- Translate languages
- Convert text to speech and speech to text
- Perform text classification
- Build a virtual agent (chatbot)
Watson is more of an assistant. It knows when to seek the answer from the knowledge base, when to ask for clarity and when to lead yourself to the human. Watson Assistant can work in any cloud-allowing businesses to bring AI to their data and apps wherever they are.
IBM Watson Assistant is marketed as a solution for companies of any size who want to build their voice or touch-enabled virtual assistant.
To create chatbot using IBM Watson API is mandatory to have a IBM/Bluemix account to start and its free (Lite Version.) Chatbot is built using intents, entities and Watson Developer Cloud to interact with the chatbot.
When we compare IBM Watson with Dialogflow, there is a question, what is better?
If you need a competent Artificial Intelligence Software product for your company you must make time to examine a wide range of alternatives. Aside from the robust features, the software which is simple and intuitive is always the better product.
In 2019, according to some market research, the user satisfaction level for IBM Watson is at 99% while for Dialogflow is at 96%. Both bot frameworks have their pros and cons. Dialogflow and Watson Assistant provide a UI tool to design conversation flow logic for complex dialogues.
Dialogflow provides maybe an easier and quicker way to create a custom conversational AI bot, while IBM Watson offering are targeting more corporations and enterprise organizations. For those who start to learn how to build a chatbot, maybe is better to choose and begin with Dialogflow.
Watson conversation is expensive compared to Dialogflow, while development interface in Dialogflow could have been better. Dialogflow bot for website integration does not support buttons and links while Watson Assistant for web integrations supports buttons and links usage.
Watson Assistant and Dialogflow integrate with variety of other popular platforms and systems.
Watson is not a single thing. Watson is a collection of APIs that can be used to solve various challenges and Watson Assistant is part of it. Many senior developers think that today there's nothing on market like Watson Assistant.
With the proper expectations and in the proper hands, Watson's APIs can be used to do some really phenomenal stuff.
More about Watson Assistant you can read at official IBM website: https://www.ibm.com/cloud/watson-assistant/
Enterprise AI Chatbot Integrations
The chatbots we create at The Bot Forge can do anything. We talk a lot about the chatbots themselves, NLP, Entities, Sentiment Analysis, Machine Learning, Training; all the good stuff which we leverage to make the optimum chat experience for our clients
However, sometimes we don't cover what goes on under the hood to ensure your chatbot does exactly what you need it to do.
Our enterprise AI Chatbot solutions' flexible nature gives you the freedom to build and expand on it however you see fit. No matter how niche your use case is, the solution will make it possible.
So what do we mean by service integrations? In this case we mean what systems do you want your chatbot to talk to or interact with to get their job done. To do it's job an enterprise AI Chatbot may need to integrate with multiple existing systems: CRM, internal knowledge base or meeting booking system. It all depends on your use-case.
Enterprise Content Management
Help your suppliers, customers, vendors and internal stakeholders in finding relevant documents quickly. Our chatbots will integrate with any internal company database or third-party database for your end-user to have appropriate human-like responses. At the same time these types of integrations allow easy management of bot responses as you have control over the single source of truth.
CRM Applications
Deliver a better experience to your consumers by integrating your customer support chatbots with Hubspot, Salesforce CRM or Zendesk.
ERP Systems
Chatbots can extend the capabilities of your ERP systems and can change the way you have been doing business.
Appointment Systems
Chatbots integrate with most popular appointment software solutions to book meeting and schedule appointments.
Check out some of the enterprise AI chatbot integrations we are using today and the potential they unlock. There are hundreds of examples, stretching across all sectors, these are just a few. Get in touch if you have your own integration in mind.
Connect chatbots to Google Sheets so your chatbot can respond accurately by connecting to up to date information.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Google Launches New Dialogflow CX in Beta
Google has beta launched a new version of it's Dialogflow natural language understanding (NLU) platform: Dialogflow CX. The Dialogflow version we all know and love is now called Dialogflow ES.
The new Dialogflow Customer Experience (CX) platform is aimed at building advanced artificial intelligence agents for enterprise-level projects at a larger and more complex scale than the standard variety.
"Dialogflow CX provides a new way of designing agents, taking a state machine approach to agent design," Google explained in the documentation for CX." This gives you clear and explicit control over a conversation, a better end-user experience, and a better development workflow."
Stand Out Features
The stand out feature for us is the that the Interactive flow visualizations allow conversation builders to quickly see, understand, and edit their work so creating more complex multi turn transactional conversational experiences will be more straightforward.
A state-based data model allows developers to reuse intents, intuitively define transitions, and handle supplemental questions.
In a single virtual agent, separate flows let multiple teams work simultaneously.
Plus, there seems to be versioning and environment at the flow level with other features such as the ability to run AB experiments and split traffic. We've started to look at the new Dialogflow CX console and things look really interesting.
A full break-down is beyond the scope of this post; we will be getting back to you with a more detailed feature analysis in future.
You can read more about Dialogflow CX here
Introductory Video can be see here
Beta Limitations
Its worth keep in mind Dialogflow CX is in beta, so some important features are not implemented yet. The following features found in Dialogflow ES are not implemented for Dialogflow CX yet:
Any language other than English (en)
Integrations
Knowledge connectors
System entity extension
History
Training data import
First Impressions
Our first impressions are that this will be a major tool for creating complex conversational enquiry heavy chatbots without having to juggle context. So particularly IVR chatbots or text chatbots which need to serve more complex roles.
It's also important to mention that this is a beta release, so some important features are not implemented yet. The following features found in Dialogflow ES are not implemented for Dialogflow CX:
As Google technology partners we are really excited about this new version of Dialogflow; if you want to learn more about Dialogflow CX and how an advanced chatbot can help your company please contact us to discuss further.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Dialogflow Mega Agent
A Mega Agent..so what?!
Ok so you could argue that I need to get out more...but I was excited to notice yesterday that there is a new feature which has sneaked into the Dialogflow console. This is the concept of a Mega Agent. It's the ability to set an agent type to mega agent so that you can combine multiple agents into one single agent.
So why is this so important? At The Bot Forge, some of our Dialogflow agents can have 1000's of intents, particularly if they are providing an information service for a knowledge base. Unfortunately, the knowledge base functionality can be limiting as looked at in my post: Dialogflow Knowledge Connectors so it's often necessary to create one intent per FAQ to get the required accuracy and control. This can quickly use up an agents 2000 intent limit.
We have recently had to look at creating our own version of a mega agent. This was to be used in a website chatbot implementation which would serve as a gatekeeper to initial enquiries so that we could hand over a conversation to a specific chatbot overseeing a specific knowledge domain. So not really ideal and involving more middleware complexity particularly as we were planning to handle some sort of context between all the agents.
There are some caveats, its still one GCP project and there is a maximum of 10 sub-agents per mega agent.
A Quick look at Mega Agents
It’s also important to remember this feature is in beta! You can read more about setting up the new Mega Agent here. At the time of writing the link on the add agent page is incorrect.
I took a really quick look at the new mega agent functionality.
Adding a mega agent
Adding a Mega Agent is pretty straightforward, when you add a new agent then you just select the switch:
Your mega agents are then listed in the agent list:
Adding a sub-agent
Once the agent is selected then a Sub Agent button is enabled:
After selecting the sub-agents button I had already created a test agent to use as my sub-agent so I connected it.
When choosing adding sub-agents you can select an environment or whether to include or exclude the knowledge Base. There is also a handy link to the sub-agent:
My test agent was a simple default agent with one added intent:
Does_mega_agent_work with one training phrase "does mega agent work"
Testing it out
So far so good. Just to recap I have created a mega agent and another agent to act as my sub. So now for a test drive of my Mega Agent in the Dialogflow simulator
Unfortunately, I didn't get the result I hoped for:
This was obviously an IAM permissions issue so I figured probably something which I had not done. I went back to the information page and re-read the section: Set up roles
Basically, to interact with a mega agent in the Dialogflow simulator, the service account that is linked to your mega agent in the Dialogflow Console needs a role with detect intent access for all sub-agents. To achieve this I went to the IAM permissions page for the sub-agent and added the mega agent's service account email address as a member of the project with a role of Dialogflow API Client.
Going back to the simulator and trying out does mega agent again resulted in the correct response from the sub-agent!
Where to go from here with mega agents.
For me, this is a major step for chatbots which have big numbers of intents > 2000. Or where different teams need to manage a particular knowledge area for one chatbot subject, use-case or topic area.
This post has really only taken a quick view of the new Dialogflow mega agent functionality. In a later post, I want to investigate leveraging contexts between agents and use a more complex example.
There are still some areas which need work though. The biggest one which springs to mind is that the training pages area of the console for a mega agent needs to be able to support the concept of sub-agents to assign sub intents. It's still just a beta feature so hopefully, more to come!
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Create a customer support chatbot for your website.
Is your business a victim of poor customer support, slow response time, and high customer demands?
You can use a website chatbot to increase customer satisfaction and retention.
Many companies suffer from the same customer support demands:
At The Bot Forge we create chatbots to be used at a variety of different customer touchpoints: Website, Facebook Messenger, WhatsApp, Slack, Microsoft teams, Alexa or Google Home.
Your customers demand faster support.
As a rule, people don’t like waiting. Keeping people waiting to get initial help can be very damaging to customer experience and directly influence customer retention.
Speed of response and speed of resolution are seen as the most important aspects of the
customer service experience, regardless of channel.
Long queue of customers waiting?
In order to help a customer effectively, your support agent can speak to at most 2 users at a time.
Many queries still come in via email which can be time consuming and often sit unanswered over long periods of time.
These types of support queries can often be handled by a website chatbot which can tap into your company knowledge base and provide support 24/7
You are spending 1000s of pounds in customer support?
You have a skilled but overstretched customer support team? Or you have outsourced it to an agency which doesn’t even understand your business or products?
You are struggleing to deal with the more complex queries because your staff are bogged down with simple customer questions which could be handled by a website chatbot.
Generate some leads whilst you provide effective customer support.
Are you looking for new ways to generate leads, turn website visitors into customers without annoying them?
In an oversaturated market, it’s best to bank on customer service as a predictor of
customer loyalty. Companies that invest in a good support not only gain through
increased loyalty and more successful upsells, but also through new customers who
are willing to pay more for a better onboarding experience.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
What Can We Expect From Conversational AI in 2021?
The use of conversational AI will continue to rise
Yes we are going to see continued growth in conversational AI in 2021. It's predicted that 1.4 billion people will use chatbots on a regular basis with $5 billion projected to be invested in chatbots by 2021. Voice assistant use will also grow. The use of voice assistants is expected to triple by 2023 (juniper research) and 50% of businesses will spend more on conversational than mobile in 2021.
Conversational commerce
For retail and e-commerce firms, conversational commerce, which is e-commerce transactions made by conversational methods such as texting and messaging, is generating waves. C-commerce not only allows brands to better serve their client base, but it’s opening up doors to new customers as well. A report made by Facebook states that 40% of global respondents said that c-commerce was their first introduction to online shopping.
97% of all respondents said that they plan to continue or increase their c-commerce spending in the future. Brands are likely to start considering how to leverage this trend and integrate messaging apps within their sales and marketing strategies.
Voice Commerce
According to Techopedia, Voice Commerce describes the utilization of voice recognition technology that enables consumers to purchase online merchandise or services.
Basically, it lets consumers buy products or services by simply using their voice. However Voice Commerce can also be part of a much wider customer journey, the transaction may not have to occur via voice. For example a consumer might have seen an ad for a product and asks Alexa about its price. The user then decides to buy it a few days later on the Amazon website. That’s why Voice Commerce involves much more than an isolated transaction process via voice.
Voice is a big deal, the number of digital voice assistants in use worldwide is estimated to reach 8 billion by 2023.
Already smart speaker users:
- research products
- add items to their shopping list
- track a package
- make a purchase
- provide ratings or reviews
- contact support
- reorder items
In 2021 there will be more Alexa in skill or Google Home action purchases as more retailers will leverage this medium; British supermarket chain Ocado has led this by example.
There will also be a continued rise in the enablement of product purchases:
Amazon has put a lot of time and effort into creating a seamless customer experience with its Echo devices.
Conversational AI taking the next steps
2021 will see a transformation for conversational AI chatbot capabilities with projects such as https://www.kuki.ai , Blenderbot , Meena and GPT-3. Open-domain chatbots will push the boundaries of what is possible. There will be an increase in AI chatbots that are personalised, processes more advanced problems and has a greater understanding of customer sentiment. In this way, your standard chatbots are likely to be replaced by conversational AI chatbots that are able to have a more human-like back and forth conversation.
These new technologies include very large language models: The Meena model has 2.6 billion parameters and is trained on 341 GB of text (1) so these models make huge computational demands. In 2021 as compute power continues to drop in price there will be a rise in availability of this open-domain chatbot technology.
Companion systems
As we live under the constraints inflicted by a global pandemic, we have been tackling an unexpected increase in alienation and loneliness in 2020. The demand to fulfil a companion role for AI assistants is something we expect to expand in 2021. With AI advancements this is becoming more realistic.
In 2020 chatbots took an informational role in many areas of the crisis; we covered a Covid support chatbot back in April. This looks set to continue in 2021.
Chatbots in immersive game experiences
Conversational AI technology looks set to be used in some really interesting ways in 2021. Particularly embedded in real-time games and integrated in multiple platforms.
Voice interaction will augment user interfaces.
We see a rise in the popularity of adding voice capabilities to software products. Specifically leveraging this sort of technology in touch screen situations. Software developers will improve their products by removing friction from the touch screen experience by bringing in voice controls. This sort of feature would be particularly useful for more complex search screens.
Conversational search
Voice search is now a rapidly growing form of access to information, but to be even more useful, it will need to become more conversational. Multiple conversational turns, follow-up on search responses, clarification and refining searches – are all aspects of natural conversation that Conversational AI is starting to replicate. These will be assisted by advancements in features such as Continued Conversation. At the same time voice search data and your own "voice" presence will become more important. Hey Google, who are "insert company name here".
Chatbots as sales assistants
In 2020 we have seen a rise in chatbots taking on the role of sales assistants. Providing specific knowledge about products is where this type of technology can excel: Providing product recommendations based on provided parameters. We've been working on these types of projects ourselves and will have more to show in 2021!
Smart IVR use will continue to grow
Speech recognition and natural language understanding for automated inbound and outbound request processing will rise. With companies offering advanced audio gateways and services such as Audiocodes. More and more legacy IVRs will be replaced with conversational IVRs: no more struggling with keypad input and overly complicated menu prompts. Advanced features such as automatic handover to live agents and multi voice options to give your Smart IVR a voice that matches your brand will improve the customer experience.
We will also see chatbot technology being utilised in different ways. Particularly in an Agent assist role, where chatbots will listen to call center conversations and provide advice, information or even responses to operatives in real-time.
Conclusion
2021 looks set to be an exciting year. Advancements in technology and changes in customer patterns and the workplace in many industries will continue to drive the growth and use of conversational AI. Everyone at The Bot Forge is looking forward to some really exciting projects in the new year!
(1) https://ai.googleblog.com/2020/01/towards-conversational-agent-that-can.html)
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
The Bot Forge Named UK Industry Leader
The Bot Forge named the leading chatbot and voice assistant agency in the UK by Clutch
Artificial Intelligence is a staple plot device in the sci-fi genre, often featured in a negative light. But in reality, predictive software improves the quality of life and leverages businesses to be more efficient and agile. No doubt AI technology can have a high-risk and high reward situation. If done correctly, its potential is unlimited. On the spot programming, automated customer support, and predictive analytics are just a few of its remarkable features.
We at The Bot Forge understand the capabilities software development and AI technology can bring to your organization. With bespoke chatbot and voice assistants as our core service, we can build the voice of your company to interact with your customers with no worries. Our process is guaranteed to make your lives, as well as your clients’ easier.
It is with great honor to announce that The Bot Forge has been chosen as one of the top AI companies in the UK. Our company is among the best on Clutch and it’s all thanks to the support of our esteemed clients.
"We are really excited to have been chosen as one of the leading chatbot and voice assistant agencies in the UK by Clutch." - Adrian Thompson, Founder of The Bot Forge
We are grateful to be recognized as an industry leader. This award and our 5-star rating wouldn’t have been possible with our clients!
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
The Bot Forge Named a Leader in AI
Did you know that more than 100,000 businesses are using chatbots to help optimize their customer experience?
Customers want instant replies, and chatbots are the way to achieve this, according to a 2018 Forbes article.
Here at The Bot Forge, we have been providing custom software development and AI services since 2018.
What they say
After working with many clients in many industries, we are thrilled to announce that Clutch, a B2B ratings and reviews firm, has listed us as one of the leading AI companies in the UK.
Additionally, we are on Clutch’s Leaders Matrix for top AI developers in the UK. The Leaders Matrix shows companies that are at the top of their targeted markets. The Bot Forge is one of the nine leaders on the Matrix.
We could not have received this recognition without our clients. We have worked with small and mid-market businesses, and these businesses represent a variety of industries.
The industries they are in include the business services, financial services, and IT industries.
We received a 5-star rating from Stitch AI, a digital engagement solutions company. We provided web development services to the company; initially, Stitch AI needed assistance in building a web portal where it could create advanced lead generation chatbots for any industry vertical.
We created a platform that helps the client manage its customers’ chatbots, and we continuously work with the client. The client has been happy with the quality of our work.
“...we’re happy with their work, and they’ve fixed any bugs in a timely manner."
— Managing Director, Stitch AI
Our Vision
At The Bot Forge, we are committed to our clients’ satisfaction. Our clients make us who we are
"Our vision is for our agency to become a global champion in creating custom chatbot solutions for our customers," said Adrian Thompson, founder of The Bot Forge.
Clutch’s sister site, The Manifest, which serves as a guide for businesses, also listed us as one of the top AI developers in the UK.
You can also see us on Visual Objects, Clutch’s portfolio-sharing sister site that features us on its list of top software developers.
Let us help your company revamp its customer experience.
Visit our Clutch profile.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Dialogflow CX Now Has a Free Trial
It was pretty exciting when Dialogflow CX was announced back in September 2020.
We talked about it briefly here.
However, Dialogflow CX can be pricey because it's based on the edition and the requests made during the month (a request being a call to the DF service via an API call or by using the console).
As a result, it's been nigh on impossible to really get to grips with it - the lack of a free tier/plan/trial, or a development version, has been a problem for anyone that isn't willing to pay to learn about the CX way of building chatbots.
And that's not a cool way to learn.
So, this month, along with some other important updates, Dialogflow has quietly announced a free trial version.
I say quietly because there was no mention of this in the usual release notes pages (normally, you can keep up to date with new announcements by following the Dialogflow release notes).
So, from now on, Dialogflow CX has a free trial - it's actually just a specific extension of the Google Cloud free trial.
Each new user will get $600 free credit to test and develop their CX chatbots.
We think that's great news!
What Else Was Announced?
New Dialogflow Messenger Integration
For us, this is a really important feature.
We love Dialogflow Messenger on ES, and we use this on a number of our chatbots deployed to websites.
Up until now, this has been missing which has been fairly restrictive. You can read more about Dialogflow and Messenger here.
CX Test Cases Launched
We think this is a really important feature and something that is vital once an agent is in production.
You can use the built-in test feature to uncover bugs and prevent regressions. You can read more about the CX test cases here.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Fred Whitton Challenge Chatbot Goes Live
The Fred Whitton Challenge consists of a 112 mile charity sportive around the Lake District. As the name suggests, it is run in honour of Fred Whitton. Fred, who died aged just 50 in 1998, was an extremely popular member of the Lakes Road Club.
Renowned for being one of the most difficult sportives it is an extremely popular event with over 2000 participants and is oversubscribed each year.
We are excited this year to have created the Fred Whitton Chatbot. On automated assistant chatbot to help the organisers.
It can answer questions and enquiries which come through Facebook Messenger providing important event information 24/7 and help event organisers answer enquires about the event, which has over 2000 participants.
The bot also allows users to look up past event times if they have participated previously and check weather reports and look up event facts and tips as well as vital safety information.
We enjoyed working on this project as the event is a well-known charity sportive. We had some great input from event organizers and coupled with the ability to train our bots to become smarter this has enabled us to answer a large number of typical event enquiries and successfully reduce the effort involved to manage them. The bot is currently able to answer over 90 separate enquiries.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
YouTube Adds Voice Search & Commands
YouTube uses voice search technology to augment it’s website user interface.
We covered this concept briefly in our What Can We Expect From Conversational AI In 2021 post.
One of our predictions for 2021 was the rise in the popularity of adding voice capabilities to software products. Specifically leveraging this sort of technology in touch screen situations. Software developers will improve their products by removing friction from the touch screen experience by bringing in voice controls.
Following this technique YouTube has added a voice commands input feature that can be used to search queries or navigate through videos on the streaming platform. Despite not being officially announced the feature is pretty useful for YouTube bingers.
Voice commands to search and play content
The new UI feature is simple to use and is very similar to the voice search function found on Android.
Voice commands can be used to scan the app for videos, navigate through results or pages, and even play content. The role also recognizes natural language commands for easy operations and supports multiple languages.
Using The Feature
We tested across 3 different browsers: Chrome, Firefox and Edge and had similar experiences. Click the microphone icon next to the search field. Once the user gives permission to access the microphone on the computer, a box appears with the word ‘listening’ within, and any video playing will pause. Clicking on the microphone button in the box will pause or restart YouTube’s listening for search terms.
The user says what they are looking for and then presented with the search results. Although not connected to Google Assistant the natural language seems to be fairly sophisticated.
Asking it to show you videos about "chatbot technology videos" will lead to a search of the correct term. However the natural language processing can still be tripped up with certain searches. For example it took a couple of tries to get the correct search for "rasa channel" to bring up the correct Rasa channel.
The search will also understand specific commands, for example If you give a command saying “play Rudimental" it finds and automatically plays a random song by the band. If you just say “Rudimental" in the voice command, it will open the official page of the band and display the list of their albums and songs.
The voice search feature can also search through your personal collections, listing watch histories and libraries, or gathering the latest videos from your subscribed channels. If asked to show the latest videos from a specific channel.
Using the voice search feature is not only limited to searching: it can be used to navigate to parts of the UI:
“Show me my subscriptions" will take you to your subscriptions list.
"show me my watch library" will take you to your watch library list.
Conclusion?.
The entire feature is essentially the same as the voice search feature added to YouTube’s mobile apps. In some ways, it’s surprising it's taken so long for voice commands to expand to the website although some browser limitations may have caused this.
Either way voice search is a useful feature to have and voice search for YouTube makes a lot of sense. The WhatsApp platform makes it easy to binge-watch endless similar videos and make it easy to randomly jump around. For kids who may struggle with spelling or other users with less dextrous fingers voice controls could be the new favourite tool. The feedback so far from our 10 year old tester and avid YouTube user is "that's pretty cool I will use that".
At The Bot Forge we feel that adding voice modality to user interfaces makes a lot of sense and a useful way of improving the usability of websites and software products.
Our own chatbot monitoring and analytics platform provides similar features.
Adding this type of feature could be beneficial to your website usability; get in touch, we would be happy to help!
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Google Launch AI Chatbot for COVID-19 Info
The Rapid Response Virtual Agent program includes open source templates for companies to add coronavirus content to their own chatbots.
Artificial intelligence and machine learning are continuing to take a front-row seat in fighting COVID-19, with Google Cloud launching an AI chatbot on Wednesday. The chatbot, which it calls the Rapid Response Virtual Agent program, will provide information to battle the COVID-19 pandemic, as announced in a Google blog.
The program will Google Cloud customers to respond more quickly to questions from their own customers about the coronavirus. It's designed for organizations who need to be able to provide information related to the COVID-19 pandemic to their customers, such as government agencies, healthcare and public health organizations, as well as travel, financial services and retail industries.
Google also offers Contact Center AI for 24/7 self-service support on COVID-19 questions via a chatbot or over the phone. Google also allows for businesses to add COVID-19 content to their own virtual agents with the ability to integrate open-source templates from organizations that have already launched similar initiatives. For instance, Verily partnered with Google Cloud to launch the Pathfinder virtual agent template for health systems and hospitals. It enables customers to create chat or voice bots that answer questions about COVID-19 symptoms and provide guidance from public health authorities such as the Centers for Disease Control and Prevention and World Health Organization (WHO), according to the Google blog.
The Contact Center AI's Rapid Response Virtual Agent program is available in any of the 23 languages supported by Dialogflow.
Google has provided a template to rapidly create a Dialogflow agent: You can find the template here. There is also documentation on how to build and deploy a virtual agent, whether voice or chat.
We've been looking in more detail at this template and created our own chatbot. This is a work in progress and will be something which we are updating and improving daily. You can interact with this chatbot in the bottom right of this page.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Why Should You Use Dialogflow For Your Chatbots?
Dialogflow is Google’s human-computer interaction developer which is based on natural language conversations. At The Bot Forge, Dialogflow is our platform of choice for chatbot construction.
There’s three main reasons for why we’re amongst companies such as Domino’s and Ticketmaster who make Dialogflow their chatbot platform of choice.
- Flexible coding: Thanks to Dialogflow’s in-line code editor, the time taken to complete code-related tasks is quicker than with other platforms. The prime benefit here is that we’re then able to spend more time perfecting the conversational experience.
- Scalability: Whether you start with 1,000 or 100,000 users, the platform can scale to your needs. As Dialogflow is hosted on the Google Cloud Platform, this allows the potential to support a user base of hundreds of millions, if required.
- Inbuilt machine learning: Arguably the biggest benefit of the platform in comparison to others is the availability of machine learning and natural language processing technologies. The access to these features allow us to create a richer and more natural conversational experience for your users. Dialogflow makes this possible by allowing us to extract data from a given conversation, in order to train our agents to understand user intents. Plus, as the technologies are already built into the platform, we’re able to construct your application much faster.
To ensure that we’re using the right platform for our clients’ needs, we continuously refresh our knowledge of other bot construction tools, such as The Microsoft Bot Framework. A benefit of using this platform from a developer’s perspective is the availability of templates to choose from, which allow for a more time efficient development. The IBM Watson Assistant is another platform that a developer may favour, as the testing the bot is simpler than it is on other competing platforms. If a priority is to feature your bot over a wide range of locations, Recast.AI may be a good option for its availability on 14 different platforms.
But, these platforms aren’t without their weaknesses. Unlike Dialogflow, Microsoft Bot Framework is lacking in the tools which help to create the “brains" of the bot, which is important for the sophistication that users are beginning to expect. Also, a downside of IBM Watson Assistant is the unintuitive relationship between intents (representation of user’s meaning) and entities (expressions recognised in categories). If you’re interested in how Dialogflow utilises intents and entities, we will be covering this in a future blog post.
Although we understand that there are features of other platforms which can make the development process more efficient, the inbuilt machine learning features of Dialogflow means we can deliver a bot that can produce a much richer conversational experience.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Why Business Chatbots Give You a Competitive Edge
Chatbots For Business
The business landscape is evolving faster and faster, we look at using chatbots for business to help you remain competitive.
There is so much coverage of artificial intelligence technology and chatbots these days There is no doubt that chatbots are big news for many different industries, from e-commerce and fashion to healthcare and banking.
Whilst many big brands have already jumped at the opportunity to leverage the technology for others it’s challenging to see where they can be a benefit to your company and whether the cost and effort involved is worthwhile. Some of you may remember many years ago when you were approached by someone selling a shiny website and then later on a new app? You probably asked yourself a similar set of questions..that’s nice and shiny but why do I want it? is it right for us? First things first let’s look at some of the basics.
The chatbot lowdown
What is a chatbot?
Briefly put a chatbot is a service, powered by natural language processing rules and artificial intelligence (AI), that you interact with via a voice or text-based chat interface. AI technology is used to enable the service to respond to specific user interaction. For example, a user could ask a chatbot a question or give it an instruction and the bot could respond or perform an action as appropriate.
This chat service can take on any number of roles, providing answers, collecting customer information, suggesting products and making sales. They can live in any major chat product (Facebook Messenger, Skype, SMS, Slack, Telegram, Viber, Twitter, Website). They can also be deployed into voice-enabled assistants such as Amazon Echo or Google home. Chatbots can also be developed to include multiple language capability.
Where can a chatbot be used?
Chatbots have been deployed in many different guises as they are extremely flexible and able to take on whatever business need arises.
You could say the possibilities are endless, here are some examples:
Celebrity
www.m.me/katyperry
Katy Perry’s official Facebook Messenger bot.
Customer Service
Vodafone TOBi
Vodafone’s customer service chatbot is based on IBM’s Watson & provides a fully integrated webchat for customers.
Productivity
AceBot
https://slack.com/apps/A0GRU84TF-ace
AceBot a productivity tool with expense tracking & intelligent task management, deployed in Slack.
Sports and Events
www.m.me/fredwhittonchallenge
The Saddleback Fred Whitton Challenge sportive bot is a smart events assistant providing event info to participants.
E-Commerce
www.m.me/LEGO
The official Lego Facebook Messenger bot. Ready to help your next LEGO purchase.
The benefits of using chatbots for your business
Provide stellar customer service 24/7
For many businesses, the biggest challenge to serving your customers in several communication channels is responding quickly all of the time.
Constantly available
One of the great benefits of a chatbot is the constant availability. Customer expectations are high expecting a quick response to enquiries. With a chatbot, you can offer your customers a service which is available 24 hours a day even when there are no employees in the office. You can rely on your bot no matter what time of the day or day of the week or timezone the enquiry is coming from.
One example from my own personal experience was with a SAAS which had charged me incorrectly for an amount of money which caused my bank account to go overdrawn. I contacted the customer support chatbot via a web interface at 1 AM and the problem was rectified and money returned promptly the next day. I went from disgruntled to a satisfied customer in a 5-minute chatbot interaction, incidentally, I’m still a customer!
It’s also worth noting that chatbots can be enabled to understand multiple languages. NLP technology will understand queries in different languages and respond appropriately. So if you support a global customer base needing to support multiple language enquiries this does not have to be a problem.
System integration
With the correct integration development, a chatbot is able to answer complex enquiries by integrating with existing CRM, ERP, CMS, and other business-critical applications.
Connect your chatbot seamlessly with your entire business ecosystem.
Scalable
Chatbots are scalable and capable of handling multiple enquiries, ready to step up when enquiry demands are at their peak.
A well implemented and executed chatbot can give businesses the ability to have more conversations and help more people at once than other alternatives, for example, live chat applications on websites.
This ability to handle the frequent enquires where the responses are often similar facilitates businesses in freeing up staff to deal with the more complex issues.
Although a chatbot cannot handle all customer queries, it can be used to deal with a large number of the routine business enquiries which most companies deal with on a day to day basis.
They improve customer satisfaction
To avoid frustration, a chatbot can be developed to use a “sentiment" function to pass users onto a real advisor if the bot can’t help or if they are not satisfied.
Other benefits can be seen in customer service gains. According to Jon Davies, head of digital at Vodafone, their customer service chatbot, TOBI provides “a far more engaging and personal" customer experience, as well as improving completion rates and reducing transaction times. These types of successes are highlighted in improved net promoter scores (NPS).
Overall chatbots for business can excel in supporting customer service teams in their communications with customers. Providing accessible information 24/7 saves businesses money and time. By 2022 chatbots are expected to save $8 billion.
Drive sales, engagement, reach
These days customers are savvier and demand an intuitive and seamless customer experience. Businesses need to consider using technology to fit in with their communication habits.
Familiar messaging technology
Many users prefer social media and mobile platforms for communication and expect businesses to be online when they are. If users are having a conversation with a chatbot in Facebook Messenger, they are using a conversation channel they are familiar with and they are already using the technology and don’t need to install a new app. The numbers of messenger app users have been steadily rising. As of April 2017 Facebook Messenger had 1.2 billion monthly active users worldwide
Use these channels to reach new and existing customers.
It’s also important to note that 2 out of 3 customers actually prefer to message a business to submit an enquiry rather than use other more traditional channels such as email or phone. Every day 1.4 billion people around the world send over 50 billion messages to communicate with each other. As messaging becomes even more central in people’s lives, demand for service in messaging has continued to rise.
The rise of voice assistants
Voice assistant technology and it’s adoption has gathered serious momentum over the past couple of years. User expectations are rising as they become educated in what it can do. As customers realise that its capabilities go beyond setting a timer, turning down the lights or playing some music; they will look to this channel to make purchases, contact customer support or use as a tool for business specific tasks.
The latest from Google Popular voice assistants currently include Apple’s Siri, Amazon’s Alexa, Google Now, Google Assistant and Microsoft’s Cortana. The big players are investing heavily in perfecting voice interfaces.
The reach of this sort of technology cannot be underestimated. You can read some of the stats and predictions for voice technology here.
Marketing clout
As an effective marketing tool chatbots can give your company an edge as they can enter into personalized and automated communication with your customers.
Using platforms such as Facebook messenger, substituting emails with push notifications can obtain much higher click-through rates. Used wisely opt-in targeted messages or push notifications have 90% read rates and a 40% click-through rate. Chatbots can be used to send users personalised tips, greetings and information, generating leads, harvesting reviews and forging stronger customer relationships.
Utilising these techniques a chatbot is able to reach participants wherever they are, regardless of where the chat session was initiated, whether on a mobile app, a website and even from social platforms such as Facebook Messenger.
Businesses are finding chatbots to be a great tool to engage with their market: “Our target customers are early adopters of social innovation so a chatbot is the perfect vehicle for us to communicate with them", Sarah Gower, Adidas.
Sales
Chatbots are ideal to answer first customer questions. if the chatbot decides that it can not effectively serve the customer, it can pass those customers to human agents. High value, responsive leads will be called by live agents increasing sales effectiveness.
Chatbots can be used to answer customer’s questions and promote products. Engage with the right customer by analyzing their profile and historical data and user characteristics. A bot can provide a channel for purchasing easily and quickly if requested.
Conclusion
I’ve really only scratched the surface of chatbot and voice interface technology capabilities and what can be achieved and how it can help your business be more competitive. However, it’s important to consider them carefully. It’s up to the business to decide if a chatbot is a right move for them, for some the business case may not be there or something to consider in the future. Building a chatbot because you think you should or because its the latest thing can only result in wasted time, money and effort.
I hope you find this post helpful in considering how using chatbots for business can help you to achieve a competitive edge.
If you already have a chatbot idea and want to look into this further have a look at our post planning the best chatbot
At the Bot Forge, we specialise in building chatbots.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Learn How AI-Powered Sports Software Helps Event Organisers
A Sports Software Chatbot Case Study: The Fred Whitton Challenge Sportive automated assistant, advanced
We report on our AI chatbot sports software project to aid the organisers of one of the UK’s most well-known cycling events.
Leverageing ai powered sports software with our core Aktivebot chatbot the goal was to create an automated assistant available 24/7 to reduce time and effort needed by event organisers to respond to event enquiries whilst still providing an easy way to contact the events team if necessary.
The Saddleback Fred Whitton Challenge is a charity event in honour of the late Fred Whitton consisting of a 112-mile charity sportive around the Lake District and is arguably one of the UK's most well known and hardest sportives with over 2000 riders and 5000 applications this year.
The Fred Whitton Challenge has been running since 1999 and as a result is extremely popular with over 4000 followers on their Facebook page where a large number of ride questions were being asked via the message me button there. We wanted the AI chatbot to assist the event organisers in answering ride and registration queries and reduce the amount of time spent answering routine questions. We also wanted to provide the ability for users to look up their time for this year and previous years.
The chatbot we created is integrated within the “Facebook Messenger app" of the Fred Whitton page and users can contact it through the private “Messages" feature of their page, or directly through the Messenger App.
The sports software project
The project brief was for The Bot Forge to create an AI powered chatbot capable of handling event enquiries 24/7 which could be deployed into the Facebook Messenger framework and utilise rich ui elements. Future deployments could be aimed at website integration.
For such a long-running event, Human Race and the Fred Whitton organisers wanted to provide the optimum user experience and still make it easy for participants to message organisers directly through the chatbot if they wanted to contact a real person by messaging them directly.
The chatbot understands human language, leveraging advanced Natural Language Processing and answers questions such as “what is the fred whitton?", “ I’ve injured myself at the weekend I need to defer till next year",“ when can I get my race pack?", “ help I need the GPS files for the route", “ Is there any way to buy a jersey post-event?","I want to contact an organiser", and “when will the results be available?" The chatbot replies to a question based on it’s own programmed data or points to the specific information on the Fred Whitton Website so that it works in tandem with the website itself.
[av_video src='https://www.youtube.com/watch?v=LUSPZnmiACI' format='16-9' width='16' height='9' av_uid='av-2b5n7a']
Press the play button to watch a real conversation with The Fred Whitton Chatbot
The technology
We used Google Dialogflow to provide the NLP engine and Google Firebase for the fulfilment hosting. The fulfilment or web-hook is where we were able to compute more complex answers for the AI chatbot to give to users and create the correct responses for. For example when looking up users past ride times, the web-hook was able to look up past results for users from a results database. Facebook ui elements added rich content, particularly useful when asked about merchandise details and availability; linking directly through to the official shop.
The conversations
The real challenge in creating the chatbot was leveraging natural language technology that can support the range of questions that event participants might ask: for example, all the different ways that people might ask about the route. We are helped in this process by our own Aktivebot pre-created sports events intents.
Small talk
The chatbot includes the ability to provide small talk, which is used to provide responses to casual conversation. This feature greatly improved user experience when talking to the agent.
Initial question data
Initially, we imported the pre-created sports events intents (an intent represents a mapping between what a user says and what action should be taken by the chatbot).
We then looked at FAQ data provided by the Fred Whitton steering committee and historical questions to their facebook page which gave us some invaluable insight. Using this information we were able to create the conversational scripts and then implement the conversation ability with each question matching an intent
This was an iterative process. Matching user intents to core functionality and features and training the natural language processor to understand users and handle conversation failure scenarios gracefully.
The conversational UI was then fine-tuned, with rich elements implemented where necessary.
What were the questions?
Most asked questions by participants match the questions that the event chatbot is able to answer, i.e.:
- Questions about registration: deferring places, available places, waiting list enquiries.
- Questions regarding merchandise: jerseys for sale on the day.
- Questions about the ride: route details, information about closed roads, clothing enquiries.
- Questions after the event: results, photos availability, the next ride date.
The training
The questions were often related to ride specific information. This meant that for an optimal intent matching rate, it was necessary to work closely with the event organisers to provide answers to specific questions. The capabilities of an ai sports software chatbot will improve over time, the more messaging transcript data the better so the more it’s used the better and more accurate it will get. Hence the training logs were checked multiple times a day and improvements made where necessary. By focusing on all questions answered it is possible to greatly improve the intent matching rate of the chatbot over time.
The training data was invaluable for perfecting the bot conversations. The process highlights any need for new responses as a continuous cycle of continuous learning.
The “training" of the chatbot can then be used from one year to the next. Any event detail changes can be carried out easily.
Results
The sports software chatbot was launched on 21st March with the scope constrained to Facebook Messenger with no advertising whilst the chatbot was evaluated.
Activity
The high number of participants using the chatbot can be explained by the fact that visitors still have questions that the website itself does not answer or does not answer quickly enough. The chatbot was, therefore, a great place to provide up to the minute event information, such as information about closed roads and the slight route change which resulted in one more hill showing.
The chatbot was not heavily advertised so we envisage activity levels will improve as participants get used to the chatbot as a resource they can use and other strategies to engage users are utilised.
The chatbot was answering questions on the run-up to the event and also during and after.
Success rate
The success rate of the chatbot to answer queries was overall around 60%. With more focused training over a longer period with another event in 2019 we expect this figure to rise until our aim of an 80% success rate is reached.
Feedback
The chatbot worked well in Facebook Messenger as its one of the preferred channels for chatbots in general. Deploying the chatbot in a chat widget as part of the website itself would undoubtedly result in more engagement and something to consider for the future.
Help intents and the handover protocol were also very successful. If a user did not get a correct response and/or wanted to get help or contact an organiser directly this worked really well. The overall feedback from users was positive. There were always some intents which the bot would struggle to match the first time which would be handled gracefully; however, due to the ability to train the chatbot, leveraging AI the correct response would be prepared for next time.
I’m impressed with the chatbot it seemed to work well. I think it is a good source of help and with it learning as it goes along it would answer lots of questions going forward. If it cannot help it still contacts the organisers where we can answer.
Carolyn Brown: Fred Whitton Challenge Steering Group — Saddleback Fred Whitton Challenge
The Fred Whitton Challenge chatbot still has many areas where it can be developed and improved, particularly by providing more integration with existing systems and utilising push notifications: this will be something carried out in the future.
Overall the success of the chatbot hightlights the benefits of deploying this type of ai sports software in sporting events and is definitely something to consider to give event organisers an advantage in a competitive market
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
The Bot Forge Featured in Milton Keynes Business Magazine
INBUSINESS COVERS OUR AI CHATBOT IN SCIENCE AND TECHNOLOGY SPOTLIGHT
It was great to have our AI Chatbot featured in the inBusiness magazine issue spotlight this month. You can read the feature here
Inbusiness is a bi-monthly publication and digital magazine created by distributed to over 3,000 business contacts in and around Milton Keynes. The June/July 2018 issue spotlight was science and technology so it was great that the editors of the magazine wanted to cover our Fred Whitton Challenge ai chatbot, particularly when the ai chatbot was created to assist the organisers of a charity ride.
You can learn more about our chatbot agency here.
We also cover further technical details about the project here.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Key Features of Conversational AI Platforms
We build a lot of different types of chatbots at The Bot Forge and deliver these to a variety of channels such as websites, Facebook Messenger, Slack, and WhatsApp.
To create our chatbots we often use different AI platforms which offer more suitable features for a specific project.
All the major cloud and open-source providers have adopted similar sets of features for their conversational AI platforms and provide good NLU (Natural Language Understanding).
There are also some strong options for open source privately hosted systems.
Conversational AI Platform Key Features
We wanted to spend some time looking at some of the more popular AI platforms in a bit more depth in this series. To help look at each one we have focused on the following specific features:
API & UI
A conversational AI platform should provide User Interface(UI) tools to plan conversational flow and help train and update the system
Context
As well as intent and entities, a context object allows the system to keep track of context discussed within the conversation, other information about the user's situation, and where the conversation is up to. This is often the NLP feature that is vital in creating a complex conversation beyond a simple FAQ bot.
Conversation Flow
Looking at the current position of a conversation, the context and the user's last utterance with intents and entities all come together as rules to manage the conversational flow.
This can be challenging to create and manage so a platforms' tools in the form of a flow engine, in code and complimented by a visual tool can provide advantages depending on the chatbot project itself.
Other features such as slot-filling (ensuring that all the entities for intent are present, and prompting the user for any that are missing) can be important.
Whilst most platforms fall into this category some systems use machine learning to learn from test conversational data and then create a probabilistic model to control flow. These systems rely on large datasets.
Pre-Built Channel Integrations
Having a conversational platform that supports your target channel out-of-the-box can substantially speed up the delivery of a chatbot solution and your flexibility in using the same conversational engine for a different integration.
This is one of the reasons we really like Dialogflow's tooling.
Chatbot Content Types
Whist the focus of a conversational AI platform is understanding pure text, messaging systems and web interfaces often involve other content, such as buttons, images, emojis, URLs and voice input/output.
The ability of a platform to support these features is important to create a rich user experience and help to manage the conversational flow.
Integrations
Bot responses can be enhanced by integrating information from the user with information from internal or external web services.
We use this type of ability a lot in creating our chatbots and in our opinion feel it's one of the most powerful features of a chatbot solution.
With this in mind, the ability to configure calls to external services from within a conversation and use responses to manage conversational flow is important in building chatbot conversations.
Pre Trained Intents & Entities
Instead of creating entity types such as dates, places or currencies for each project some systems provide these pre-trained to deal with complex variations.
In the same way, common user intents and utterances such as small-talk are offered pre-trained from some platforms.
Analytics & Logs
The key to creating a successful chatbot is that they need to be constantly trained and monitored.
To aid in continuously improving the system once initially launched, the conversational tools should provide a dashboard of the user conversations; showing stats for responses, user interactions and other metrics. Export of these logs is also useful to import into other systems.
Other important AI features enable easily training missed intents, catching bad sentiments and monitoring flow.
Tech Stack
It can be important to take into account what libraries are provided by an AI platform and in what supported languages.
In the end, the stack may favour a particular platform if it fitted with your current codebase or teams skillset.
However, as a full-stack javascript software house, we find Node.JS to be our server stack of choice when building our bots and most AI platforms cater for this.
Costs
These are the costs for cloud hosting and cloud NLU solutions. An important aspect to consider particularly for large scale enterprise chatbots handling large volumes of traffic where NLU monthly costs can reach £thousands.
Many providers offer a free tier for their AI platform solutions. A paid-for tier will then normally offer enhanced versions of the service with enterprise-focused features and support for greater volume and performance.
Costs tend to be charged in one of 3 ways, per API call, per conversation or daily active user and also per active monthly user (normally subscriptions are in tiers).
We try and look at costs as publicly published for the paid-for plans suitable for enterprise use in a shared public cloud environment.
The Platforms
Keeping all these feature sets in mind we hope to look at the following AI platforms over the coming posts.
- Botkit
- Chatfuel
- Amazon Lex
- Microsoft Luis
- Google Dialogflow
- Rasa
- IBM Watson
Please get in touch if you feel we should look at a platform that we have missed!
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Dialogflow Enterprise Edition Announced
It was exciting to hear that Google has announced the beta launch of its enterprise edition of Dialogflow, its tool for building chatbots and other conversational applications.
What Does This Mean For Our Customers?
It comes with a number of benefits, including:
- Basic analytics and monitoring capabilities
- Built-in support for speech recognition
- 24/7 support
- SLAs and enterprise-level terms of service promising data protection
- Higher text query quota
- Now part of Google Cloud
The new enterprise launch is an important addition, meaning The Bot Forge can provide an improved service to customers looking for an enterprise-grade chatbot.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Building a Chatbot Using Amazon Lex
What is Amazon Lex
Amazon Lex is a service by AWS for building conversational interfaces into any application using voice and text. Lex has quickly become popular among chatbots enthusiasts looking to leverage the technology which powers Alexa.
Users can be up and running within minutes with no upfront costs.
Amazon Lex has been in development since 2010 but was first made available in 2016. The system consists of the technology that powers the automatic speech recognition (ASR) and natural language understanding (NUR) capabilities of the Alexa console.
Amazon Lex is a fantastic choice for building chatbots. The main selling points are the system's ease of use, pay-as-you-go pricing, and excellent opportunities for integration over the cloud.
Artificial intelligence capabilities
Amazon Lex is powered by the same engine as Amazon Alexa. As a result, the system ranks highly in terms of speech recognition, and complex nuances, and sophisticated language understanding.
Developers can use Amazon Lex to create chatbots for voice and text which can be employed for a variety of uses, including customer service, taking orders/reservations, or controlling IoT devices.
Amazon Lex can begin building conversational interfaces with just a few example phrases. With intent chaining, the developed chatbot can suggest the next topic and switch dynamically.
Lex also supports advanced features such as Slot filling and you can meet pretty much any integration requirements by leveraging Lamba functions.
Admin platform
You must first have an Amazon Web Services account before using Amazon Lex. Access is found through the AWS Management Console.
Users can then use the Amazon Lex console to create and deploy speech or text chatbots directly to new or existing chat applications, web apps, and mobile devices like Slack, Kik, Facebook Messenger, or Twilio SMS.
Fully managed and scalable
One of the best things about Amazon Lex is that it is fully managed, so as your user base grows you don't have to worry about hardware or infrastructure. The system also offers an unparalleled opportunity for integration with other services.
Through AWS Lambda, Amazon Lex supports enterprise integrations. With Lambda, you can run code for virtually any type of application or backend service.
So integrating with CRM, ERP, Appointment systems, or content management systems is all achievable whatever the use-case.
Pricing
Amazon Lex has no minimum fees or upfront costs. Users are charged on a text or speech request basis, so you'll only be charged as much as you use it.
Lex is also a pay as you go service, so there is no recurring fees or subscription. You can also test the system and build your initial chatbot at no cost whatsoever.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
6 Common Mistakes to Avoid When Developing a Chatbot
We are experts in developing chatbots so we know If you are looking to streamline certain operations of your business, developing a chatbot is a great way to go about it.
After all, you can use technologies such as artificial intelligence (AI) and natural language processing (NLP) so that it can follow different types of conversations with users and provide relevant responses where necessary.
Chatbots are no longer a gimmicky tool available on the internet, as they have gained popularity and sophistication many users are incorporating them into their digital strategy. For example, a large number of businesses now use bots as part of their customer service. As these chatbots never go offline, they are always available to assist users, at any given time of day.
It is a lot cheaper for a business to implement a bot as part of its customer service than hire employees. If you employ an individual, you have to train the person and provide a salary and vacation time. One example is one of our clients https://amicable.io who replaced call centre resource with a chatbot to book client meetings.
Chatbots have a lot of advantages, which explains why businesses want to make the most of them. However, while creating these bots, it is natural to make errors, which hampers the user experience. As this might be the first or nth time you are developing a chatbot, you want to make sure it functions as expected. Here are six common mistakes to avoid along with how you can overcome them:
1. Assuming every user wants to talk
You tend to believe that everyone who visits your page or installs the chatbot on Slack, or other popular messaging platforms wish to start talking to it immediately. However, most of the people on the internet don’t want to communicate with the bot, unless it is necessary.
One reason why chatbots are great marketing tools is that they can engage with prospects by answering important questions. As a result, it brings down the sales friction, making it simpler for the user to invest in what you have to offer.
If your chatbot starts to message the individual as soon as he/she opens it, there is a high chance the person will find it annoying. A better practice would be to wait for the user to respond or you can leave instructions in the description on how to start conversing with the chatbot. Our sports events Facebook Messenger chatbot Carly utilises this kind of functionality enabling users to set push notifications for their any new sports events based on their own criteria.
2. Developing a chatbot but failing to track it's performance
So you've developed a chatbot but since the chatbot makes use of the latest technological advancements in the industry, you might assume that you shouldn’t keep an eye on it's performance. After all, you spent a considerable portion of your time training it, so that it can have a continuous conversation with your customers.
However, you will never know the effectiveness of your bot, if you don’t track the key performance indicators (KPI). These metrics provide a deeper insight into how you can continue to improve your chatbot. For example, you can see where most of the users tend to leave the conversation. With this data, you can think of different ways to keep them engaged so that they continue to talk to your bot. We feel that the history and training tools provided by Dialogflow enable us to track chatbot performance effectively.
3. Forgetting to list in directories
Once you have completed developing your chatbot and its up and running on various messaging platforms or your website, its a mistake to think you completed your job. All your visitors have to do is start talking to the bot, and it will help them in their tasks.
However, not everyone will know about the existence of your chatbot. Several messaging platforms may not have powerful search, which makes it harder to discover your bot. The best practice is to find third-party websites and lists your chatbot in it. As a result, if people look for your bot on Google or other search engines, the chances of it popping up in the first page of results goes up significantly. The best place to market your own new bot is on your website, why not write a blog post about your chatbot journey, you can guarantee other companies will be interested in your journey.
4. Impersonal conversations
The reason why people don’t like talking to bots is that the conversation tends to be boring and bland. As a result, they prefer to converse with human beings, as the experience is better in every way.
Think about it, would you like talking to a chatbot which sounds like it is speaking in a monotone? Rather than putting your bot in the same position, you should think of different ways to spice up the conversation. For example, you can ask the user what the chatbot should call the individual while talking to one another.
One thing is key here and we have seen this in our experience: to gain better customer satisfaction it's better to explain to your users that your bot is a chatbot and not try and masquerade as a human. If users are aware they are talking to a chatbot from the off it will gain confidence and improve the customer experience as the user becomes more forgiving.
5. Not paying attention to a chatbot's persona and tone.
Since the entire conversation between the chatbot and its users is going to take place via text, you need to pay close attention to it's tone. Deciding on a persona for your chatbot is part of the conversational design process. Using the right type of communication will determine whether your bot performs well among its intended target audience.
While this tends to be challenging, there are several ways you can overcome this obstacle. For example, you can ask a small number of people from your target audience, what tone they would find appropriate. At the same time, you can also have a beta group, which allows you to experiment and see which one works well. Matching tone to the industry and subject matter is important to build a satisfying experience for your chatbot users.
6. Help features and live agent handover
Since the purpose of the chatbot is to reduce the workload of your employees, you tend to assume that you don’t need a live agent. The problem with testing is that it may not take into account all the variables present in real-world scenarios. As a result, when you deploy your chatbot, it might not know how to handle a specific question.
Due to this reason, it can go on an endless loop, and the only way out of the conversation is to quit or restart the chatbot. An excellent way to overcome this problem is to allow your chatbot to ask a live agent to join the conversation during this situation. Once the employee helps out the user, he/she can provide information to the developers on how to improve the communication skills of the bot.
It's also important to provide easy help for users to access during the bot conversation. At The Bot Forge we always implement a help feature for our chatbots so users know what they can do and how they get back on track our Facebook Messenger chatbot for the Fred Whitton Challenge is a perfect example.
Chatbots are becoming a great tool for businesses. You can use them to make life easier for your clients, by assisting them in various functions. By knowing what the common mistakes are, you can avoid them entirely and design the best bots in the industry
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Discover The Benefits of a WhatsApp Chatbot
Creating a WhatsApp Chatbot with Twilio & Dialogflow:
In this article, I’m going to cover WhatsApp business and dive into creating a WhatsApp chatbot.
Why is WhatsApp So Important?
It's official, WhatsApp has one or two users… yes, that’s the understatement of the year!
In fact, WhatsApp has 1.5 billion users from 180 countries, which makes it the most popular instant messaging app worldwide. This messenger is handy for being secure, fast, and easy to use. It’s not just about the massive number of users though, it’s about engagement. WhatsApp users send about 65 billion messages per day, that is about 750,000 messages per second! WhatsApp usage shows no signs of slowing down.
WhatsApp for Business
In 2018 WhatsApp announced the official launch of their platform made for business.
This allows companies to communicate with clients using WhatsApp for Business instead of having to use their own personal numbers.
This will allow companies to automate, sort, and respond to messages on this incredibly popular messaging channel.
As of May 2018 WhatsApp for business had 3 million users.
Whatsapp is well known for protecting your data which includes chats, documents, status updates, photos, videos, voice messages, and calls via WhatsApp’s end-to-end encryption.
As a customer, you know you’re interacting with an officially approved business and all of your rights are protected by WhatsApp’s secure environment so its no doubt that the Whatsapp business client offering is becoming increasingly popular.
Whatsapp Business App
Business Profile
The WhatsApp business client has been built with the SME in mind. The app can help you provide customer support and deliver important notifications to your customers. These WhatsApp Business accounts help brands to improve brand loyalty. A business profile gives the company a familiar “face" and identity.
First off you need to grab the Whatsapp Business App for your mobile phone of choice which is free to download.
Users can create a business profile with helpful information for their customers such as their address, business description, email address, and website.
Steps - Update your business: Open the Whatsapp Business app → Open Settings → Open Business Setting → Business Profile.
Messaging Tools
The Business client provides some really useful automated messaging functionality.
Welcome Message
You can tailor your own greeting message and send to customers who message you for the first time or after 14 days of inactivity.
Quick Replies
Businesses can create their own standard quick reply messages to streamline their conversations and save time.
Away Message
You can tailor your own away message and reply when you are away.
Steps - Use messaging tools: Open the Whatsapp Business app → Open Settings → Open Business Setting → Select Away message/Greeting message/Quick replies.
Contact Labels
Another useful feature is the ability to organise chats and contacts with labels.
Steps - Use Labels: Open the Whatsapp Business app → Open Chat → Open Menu → Select Label chat
Statistics
The business app also provides statistics covering messages sent, delivered, read, received.
Steps - Access Statistics: Open the Whatsapp Business app → Open Settings → Open Business Setting → Statistics.
WhatsApp Business API
WhatsApp Business API is the enterprise offering for the platform.
Prerequisites
The prerequisites for using WhatsApp commercially via the WhatsApp Business API is to either apply for an own account directly from WhatsApp or to buy access from one of the official Solution Providers.
Access to the WhatsApp API has been limited, to say the least, the program is currently in a limited public preview, In fact, at the time of writing, there are only around 40+ companies listed as solution providers. You can still request access but there is no guarantee when/if this will be provided, I think Facebook will be favouring end client/solution provider applications with large estimated numbers of messages. Once you have gained access you will also have the technical challenges of getting set up. A quicker/simpler option is to use one of the solution providers for now. At least whilst you wait for your application access to be approved!
For the purposes of this article, we are going to look at using Twilio as our WhatsApp solution provider.
What's a WhatsApp Chatbot & What Are its Benefits?
A WhatsApp chatbot is similar to a Facebook Messenger Chatbot. When a user interacts with (WhatsApps) your number then the response is handled by your chatbot.
So what are the benefits?
- Customers can contact your business on their preferred platform, which they use daily
- It supports the ability to respond to customers questions right away
- You can reliably send mission-critical messages from delivery notifications to booking confirmations and delivery alerts
- Leads customers down the sales funnel by enabling them to take fast actions
- It builds trust and loyalty with customers
- Personalization of customer experience is possible by customizing the script that WhatsApp uses
- Customer communications are secure with an end to end encryption in WhatsApp
- You can send images, audio, video and pdf files via WhatsApp
Creating a WhatsApp Bike Shop Chatbot with Twilio & Dialogflow
We are going to look at building a WhatsApp chatbot for a bike shop.
The chatbot will be able to answer a simple set of bike shop related questions and book your bike in for a service.
We will use Dialogflow to create our chatbot and then connect this to the Twilio Sandbox for WhatsApp. The sandbox enables us to prototype with WhatsApp immediately, without waiting for the approval of our number.
There are some things to consider using the sandbox:
- You can only message users who have joined your sandbox. Messaging other users will fail.
- Load testing profile traffic is not supported
- The Sandbox numbers are restricted to 1 message every 3 seconds
- Sandbox numbers are branded as Twilio numbers
- You can only use pre-registered templates with the sandbox for outbound messages sent outside a WhatsApp session.
If you tell your customers that you will be using their email address and mobile phone number to send them information about your services and products, you should do that and nothing more.
Step1: Set Up Your Twilio Account
Signing up with Twilio is the next step and it's free with no need to provide a credit card, bonus!
- After opening the Twilio website, click the “Try the Sandbox today" button
- Sign up for Twilio account
- You will need to authenticate your email address.
- It's also useful to fill out the customisation clarification questions. “ Welcome! Let’s customize your experience!" However, it's not essential and you can just skip to the dashboard. We answered the questions and selected Send WhatsApp messages when asked: “ What do you want to do first?"
- Go to the Products section, then Programmable SMS and then navigate to the Twilio console
- You are then asked to activate your sandbox and agree to the WhatsApp terms of service.
- Go through the steps to set up your testing sandbox in the learn section.
- Try sending messages:
1. First send a message to the test number to link your number to the sandbox.
2. Then Send a One-Way WhatsApp Message. It's interesting to note that you must use a pre-approved template from WhatsApp.
3. Try sending Two-Way Messaging. Note 2-way messaging means you now have 24-hours between your Sandbox and your WhatsApp account, without the use of templates.
Step 2: Create your Dialogflow Agent
We won't go into detail here into how to create a Dialogflow agent you can learn more here there are plenty of good resources, we recommend taking a look at this.
If you want to use the agent we have built you can create an agent and use the restore from zip feature of the Dialogflow console to import our agent which you can download from here.
Step 3: Enable Twilio Integration in Dialogflow Agent
In the Dialogflow console → Under integrations → select Twilio (Text messaging) → in the settings window, there will be a Request URL (seen here in green).
Copy this URL and go to your Twilio account in the Sandbox configuration and paste into the “WHEN A MESSAGE COMES IN FIELD".
Once you’ve done that go back to your Dialogflow Twilio settings window and input the rest of the account details: Make sure you have your Twilio API Credentials to hand, you will need Account SID, Auth Token, Phone number — Used to authenticate REST API requests.
Steps- API Credentials: Log-into Twilio → Open Dashboard → Open Settings → General.
Step 4: Test your Agent
At this point, if you have properly carried out all the steps you should be able to send a WhatsApp message to your number and the response will come from the Dialogflow chatbot. You can see ours below. Notice the sandbox limitation; Sandbox numbers are branded as Twilio numbers.
Adding Other Cool Functionality
There are loads of other cool features we could add to our Bike Shop WhatsApp chatbot, for example:
Use the Twilio WhatsApp API to send customers a notification that their bike is ready! You can read more from the API Reference here.
The WhatsApp message can be sent using a pre-provisioned template e.g.
Hi {{1}} your bike {{2}} is completed and can be collected when it's convenient, the cost for the work is {{3}}. Details of work carried out: {{4}}
Once your Twilio number has been enabled for WhatsApp you can also create your own templates.
Going into Production?
If you want to start using the Twilio API in production you need to enable your Twilio numbers for WhatsApp.
This involves initiating a request via Twilio.
Fill out this form to send the request. Once you have provisioned your numbers you also need to provide Twilio with your Facebook Business Manager ID.
Creating a WhatsApp Chatbot with Trengo
Trengo offers an omnichannel collaborative platform for its customers.
They provide the ability to support enquiries across multi-channels.
One of the cool things about their technology is that their platform supports Business WhatsApp and they have the ability to create chatbots on their platform which can be directly linked to a Dialogflow agent, which is great if you want to create a Dialogflow powered chatbot and easily connect to a WhatsApp number!
Conclusion
WhatsApp is a platform that connects billions of users every day and is now granting businesses endless possibilities for reaching and engaging with their massive audience.
Using WhatsApp for business, companies are now able to interact with customers on the platform that they love and already use.
Hopefully, you’ve enjoyed following the article and you can see the potential for using WhatsApp chatbots in your business.
We looked primarily at Twilio as the WhatsApp API provider, however, there are other providers which we will be looking to cover in the future, here are some of them:
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
A Guide to Conversational IVR
So what is conversational IVR (Interactive Voice Response) and why should businesses care.
According to Forrester Research, customers expect easy and effective customer service that builds positive emotional connections every time they interact with a brand or organization.
For businesses improving their organization’s customer experience is a high priority.
Additionally, 40% of surveyed business leaders say that improving their organization’s customer experience (CX) is a high priority, ahead of initiatives like improving products and differentiation and reducing costs.
Despite the growth of customer service via chat and email, dialing a phone number is sometimes the best option for clients to make initial contact, resolve issues, and receive customer support. Whilst for some use-cases such as the university clearing process, it’s the norm.
Unfortunately, most outdated legacy interactive voice response (IVR) systems were never designed with CX in mind and unable to handle nuanced customer enquiries which a menu system is not designed to address. At the same time, some businesses can’t afford the staff to take phone calls 24/7 or provide enough capacity to manage spikes in demand.
What exactly is Conversational IVR?
Conversational IVR is a software system which uses voice commands from customers to allow them to interact with IVR systems over telephony channels.
The development of voice gateways has enabled modern chatbots to be connected to pre-existing telephony engagement channels via SIP/RTP and as a result massively extended the reach of chatbot technology into contact centres, enquiry numbers and helpdesks across all industries.
Voice gateways provide the technology to connect Telephony services over SIP to chatbots. For example Audiocodes PNC Voice.AI Gateway.
Bringing natural language processing(NLP) to standard telephony systems means both intent and context can be understood by these systems. It removes the burden on customers to navigate through slow, confusing and hierarchical menus and simple voice implementations of legacy IVR. And it lets customers self‑serve and resolve issues within the IVR System.
Conversational IVR takes auto-attendants and IVRs to the next level providing human like experience by enabling more human-like multi-turn interactions, leveraging natural language processing, artificial intelligence, and machine learning.
The ability to act on enquiries by extracting intent and variables from a conversation means that a conversational IVR system has the potential to work through customer enquiries, field/re-route calls effectively and enrich customer support over telephony channels.
What’s the difference between conversational IVR and legacy IVR?
The IVR systems have been around for nearly five decades now. The technology’s commercial application was rolled out in 1973 and has 1000s of present-day implementations.
In its simple guise, IVRs are a touch-tone input and voice output system. Pre-recorded messages prompted callers to put in their request by pressing keys on a phone. Unfortunately although advanced at the time, the large menus, numerous options, and using the same handset to key in inputs and listen to the voice prompts makes the experience cumbersome and not really fit for purpose when providing modern customer support.
Traditional automated call center solutions ultimately depend on number selections or similarly basic input from the customer and have minimal ability to adapt; particularly annoying when someone makes a mistake with their input such as pressing one wrong button while entering a long account number. Accessing the correct information can often be frustrating with nested menus.
Whereas traditional IVR systems had speech recognition technology to handle simple voice commands such as “yes" or “no," conversational IVR allows people to communicate their inquiries in more complete phrases via a natural language understanding. Callers can describe questions or concerns in their own words which are then matched to intent by natural language understanding.
Leveraging machine learning to improve the NLP capability, therefore, allows modern natural language processing systems to be trained to understand 1000s of different intents (questions) with each one of these intents being asked in 100s of different ways. This also means that smart IVRs can continue to improve whilst still handling failures gracefully. For example, if a caller says something which the system does not understand, it can redirect to a live agent via intelligent call routing or instead ask further qualification questions such as requesting a customer to spell out their specific details. The next time the smart IVR encounters this question(utterance) it will have learned from past experience and fully understand this enquiry.
Systems can also leverage advanced metrics such as sentiment to streamline conversations and improve customer experience with a more human-like interaction.
The best conversational IVRs enable a more free, human-like service experience for customers, who aren’t bound to a specific menu script. These smart IVRs are more capable of guiding customers toward self-service-type solutions instead of involving a live agent. Not only does this maximizes employee productivity, handle time, money saved on staffing costs, but in a best-case scenario, it makes for happier customers who get fast, personalized responses from an automated system that isn’t entirely stripped of the human touch.
Is conversational IVR right for your business?
Bigger organisations fielding large volumes of calls are likely to benefit the most from conversational IVR. However, if you experience spikes in demand and your agent’s struggle with timely answers to some of the calls they handle in peak times then a conversational IVR system may bring a ROI sooner than you expect.
Anywhere that there is repeat demand for a specific customer enquiry could mean that the intelligent use of conversational IVR could remove a business pain point, impacting call center and customer service staff positively.
Due to the flexibility of smart IVR systems, it’s often easy to create a proof of concept implementation to validate and test specific use-cases before investing fully in a system.
Voice gateway technology also enables modern IVR to connect easily with current legacy systems.
What are the benefits of using conversational IVR
Its important to understand how conversational IVR can be leveraged to cut budgets, increase efficiency, improve customer satisfaction and meet spikes in demand.
Serve your customers faster with more precision
Conversational IVR reduces the amount of time needed to support each client. It often decreases the time taken to serve customers as users are often able to request their intention with one sentence, rather than navigate through confusing menu systems.
Leveraging natural language and machine learning to improve responses therefore enables Conversational IVR systems to always be improving and still be flexible enough to be easily changed to meet new questions and offer more detail where needed.
Cut costs
Its difficult to calculate the average cost of a live agent customer service phone call due to the number of variables. However there is no denying the need. In one report, IBM reported that worldwide companies spend over $1.3 trillion to serve 265 billion customer calls each year.
In a report by Forrester¹ on the total economic impact of smart IVR technology the projected return on investment (PROI) was judged to be between 103% and 291% over 3 years. Most organisations would benefit from using conversational IVR to satisfy customers without using a human agent. Financial gains are seen from:
- Reducing the number of calls reaching human agents by improved initial resolution rates and containment.
- Improving agent efficiency with augmented technology such as agent assist and predictive analytics.
- Improving customer satisfaction and agent experience by reducing the burden on agents so they can focus on addressing each customer’s specific request or need.
¹ New Technology: The Projected Total Economic Impact™ Of Google Cloud Contact Center AI 2020
Improve customer satisfaction
It’s no surprise that historical feeling towards automated phone services was negative due to the poor customer experience from legacy IVR systems. Overall voice was often judged suspiciously. These days opinions are shifting toward acceptance amid the rising adoption of personal assistants such as Google Assistant, Siri, Cortana, and Alexa on smart device and mobile. People are increasingly familiar with these technologies, the ways they can be used, and their limitations. Customer satisfaction numbers for popular voice-controlled assistants are as high as 80% or better depending on the platform and survey.
Users realise that they can get more done with voice these days. At the same time voice technology and capabilities such as agent assist can also empower agents to provide better service. With the careful implementation and design of voice assistants it’s possible to achieve higher levels of customer care and improved efficiency and as a result provide a better service for your customers.
Conversational IVR Provides Better Customer Support
Customer satisfaction is a critical concept for customer success professionals to understand and live by, and it’s actually about more than a money-back guarantee.
Conversational IVR modernizes conventional IVR principles with innovations such as AI and machine learning. Instead of navigating push-button menu flows, users can provide spoken inquiries reducing the friction and poor user experience of legacy systems. Smart IVR has the ability to provide responses to complex enquiries and give a response in real-time within seconds. Tasks that used to require a lengthy phone call can now often be done quickly and easily.
At the same time technology can be leveraged to provide streamlined features: sentiment analysis, call routing, fall-back handling, even energy detection tracking to provide better support. And if the customer needs to speak to someone or requests to do so, we can connect them to an agent seamlessly and retain all the information already shared in the chat.
The end result is that this allows conversational IVR systems to complete requests faster via customer self-service options in real-time. Along with maximizing efficiency and helping to offset spikes in call volume, call center costs are reduced by lowering customer churn, boosting brand perception, and improving client retention.
How to Train a Chatbot
Training data for chatbots.
I'm going to look at the challenges in creating a chatbot which can answer questions about its specific domain effectively. In particular, I'm going to look at the challenges and possible solutions in creating a chatbot with a reasonable conversational ability at their initial implementation. Every chatbot project is different but often clients come to us with a large knowledge base which they want a chatbot to support from its release but with very little training data.
We are going to concentrate on a Dialogflow project to look at some examples however the challenges and solution are similar for all the most well know NLP engines, Watson, Rasa, Luis etc.
The Challenge
One of the key problems with modern chatbot generation is that they need large amounts of chatbot training data.
If you want your chatbot to understand a specific intention, you need to provide it with a large number of phrases that convey that intention. In a Dialogflow agent, these training phrases are called utterances and Dialogflow stipulate at least 10 training phrases to each intent.
Depending on the field of application for the chatbot, thousands of inquiries in a specific subject area can be required to make it ready for use with each one of these lines of enquiry needing multiple training phrases.
The training process of an ai powered chatbot means that chatbots learn from each new inquiry. The more requests a chatbot has processed, the better trained it is. The NLU(Natural Language Understanding) is continually improved, and the bot’s detection patterns are refined. Unfortunately, a large number of additional queries are necessary to optimize the bot, working towards the goal of reaching a recognition rate approaching 90-100% often means a long bedding in process of several months.
Data Scarcity
One of the main issues in today's chatbots generation is that large amounts of training information are required to match the challenges described previously. You have to give it a large number of phrases that convey your purpose if you want your chatbot to understand a specific intention.
To date, these large training corpus had to be manually generated. This can be a time-consuming job with an associated increase in the cost of the project. One of the main issues we have faced is that often clients want to see quick results in a chatbot implementation. These types of chatbot projects are often use cases which are providing information regarding a wide-ranging domain and may not necessarily have a lot of chat transcripts or emails to work with to create the initial training model. In these cases there is often not enough training data and so it takes time to get decent and accurate match rates.
The Solution
THE BOT FORGE PROVIDES
CHATBOT TRAINING DATA
CREATION SERVICES
The Bot Forge offers an artificial training data service to automate training phrase creation for your specific domain or chatbot use-case. Our process will automatically generate intent variation datasets that cover all of the different ways that users from different demographic groups might call the same intent which can be used as the base training for your chatbot.
Multi NLP platform support
Multi-language support
Our training data is not restricted solely to Dialogflow agents, the output data can be formatted for the following agent types:
- rasa: Rasa JSON format
- luis: LUIS JSON format
- witai: Wit.ai JSON format
- watson: Watson JSON format
- lex: Lex JSON format
- dialogflow: Dialogflow JSON format
We provide training datasets in 100+ languages
We offer our synthetic training data creation services to our chatbot clients. However, if you already have your own chatbot project and just want to boost its conversational ability we can provide synthetic training data to meet your needs.
Testing the Solution
We wanted to test the effectiveness of using our synthetic training data in a Dialogflow chatbot agent by varying the number of utterances per intent using our own synthetic training data.
Dialogflow test agents
We carried out three different tests (A B and C) with 3 separate Dialogflow agents. Each agent had identical agent settings. The agents had 3 identical intents to provide information about the topic of angel investors:
- what_is_an_angel_investor
- what_percentage_do_angel_investors_want
- do_angel_investors_seek_control
In the first test (A) the chatbot was trained with 2 hand-tagged training phrases (utterances) per intent. Test (B) had 10 training phrases from our own synthetic training data per intent and test (C) had between 25 and 60 training phrases per intent.
The Test
We tested each agent with 12 separate questions similar to but distinct from the ones in the training sets.
We didn't carry out any training during testing once the chatbots were created.
We recorded the % of queries matched to the correct intent, the incorrect intent or no match and also the intent detection confidence 0.0 (completely uncertain) to 1.0 (completely certain) from the agent response.
Overall test results
|
|
% correct match rate
|% incorrect match||
|
%no match
|
|
Average Intent Detection Confidence
|Test A (2x utterances)||50%||42%||8%||0.6437837225|
|Test B (10x utterances)||91%||9%||0%||0.7590197883|
|Test C (25-60x utterances)||100%||0%||0%||0.856748325|
Test A provided a 50% match rate. We observed a significant improvement in test B with the introduction of some of our synthetic training data to the agent. We were able to improve the match rate from 41% to 91% whilst TestC with 25-60 training phrases enabled a match rate of 100%. The average intent detection confidence also grew
In summary, chatbots need a decent amount of training data to provide accurate results. If there is not enough training data then a chatbots accuracy is affected and it can take some time to train it whilst being used to reach acceptable performance levels. At the same time, it can be costly and time-consuming to create training data for a chatbot needing to handle large numbers of intents.
Our synthetic training data creation service allows us to create big training sets with no effort thus reducing initial costs in chatbot creation and improving the usability of a chatbot from the initial release stages. If you only have a limited number of training phrases per intent and have large numbers of intents, our service is able to generate the rest of variants needed to go from really poor results to a chatbot with greater levels of accuracy in providing responses. We have carried out these tests with Dialogflow, but our conclusions are relevant for ML-based bot platforms in general. We can conclude that our Artificial Training Data service is able to drastically improve the results of chatbot platforms that are highly dependent on training data
Chatbot Training Never Ends!
I've looked at the benefits of using our training data at the early stages of a chatbot project. However, it's important to note that the key to success, in the long run, is to constantly monitor your chatbot and continue training to get smarter. Either by doing constant training with human effort or by scheduling regular training cycles, incorporating new utterances and conversations from real users.
If you want to know more about our chatbot training data creation services get in touch
Appendix
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
The Non-Technical Guide to Popular Conversational AI Terminology
Conversational AI Terminology Cheatsheet
Conversational AI technology is not new, but the advanced in the technology has driven a major growth in the industry and what can be achieved in its role solving business problems for many types of industries.
We talk about Conversational AI a lot on our website and blog, after all this technology is at the core of what we do at The Bot Forge.
You may well have encountered some of the different terminology used. But what do developers and technologists really mean when they use these terms? Having a simple understanding of some of the more frequently used terms can be useful when thinking and talking about your chatbot or voice assistant strategy. This conversational AI terminology cheatsheet aims to help you understand; no technical knowledge required!
-
Algorithm
An algorithm is a formula for completing a task. Wikipedia states that an algorithm “is a step-by-step procedure for calculations. Algorithms are used for calculating, automated processing and data processing and provide the foundations for artificial intelligence technology.
-
Artificial Neural Network
Artificial Neural Networks or ANN are artificial replicas of the biological networks in our brain and are a type of machine learning. Although nowhere near as powerful as our own brains they can still perform complex tasks such as playing chess, for example AlphaZero, the game playing AI created by Google.
-
Artificial Intelligence
AI research and development aims to enable computers to make decisions and solve problems. The term is actually a field of computer science and is used to describe any part of AI technology of which there are 3 main distinctions (1)
-
Big Data
Big data describes the large volume of data – both structured and unstructured – that floods through a business and its processes on a day-to-day basis. In the context of AI big data is the fuel which is processed to provide inputs for surfacing patterns and making predictions.
-
Chatbots
I think we have mentioned these once or twice! A chatbot is a conversational interface powered by AI and specifically NLP. They can be text-based, living in apps such as Facebook Messenger or their interface can use voice-enabled technology such as Amazon Alexa.
-
Cognitive
Cognitive computing mimics the way the human brain thinks by making use of machine learning techniques. As researchers move closer towards transformative artificial intelligence, cognitive will become increasingly relevant.
-
Conversational Design/Conversational Designer
Whilst not a technical term its a relatively new role which has grown to being a vital one with the rise in the popularity of conversational experiences. It's important to understand what this new breed of skilled professional brings to a chatbot project and why they are so important. Conversation design is the art of teaching computers to communicate the way humans do. It’s an area that requires knowledge of UX design, psychology, audio design, linguistics, and copywriting. All of that put together helps chatbot designers create natural conversations that guarantee a good user experience.
-
Deep Learning
Also known as a deep neural network, deep learning uses algorithms to understand data and datasets. Deep Learning is a subfield of machine learning concerned with algorithms inspired by the structure and function of the brain called artificial neural networks. Deep Learning techniques have become popular in solving traditional Natural Language Processing problems like Sentiment Analysis.
-
Entity and Entity Extraction
Entities are also sometimes referred to as slots. An entity is used for extracting parameter values from natural language inputs. Any important data you want to get from a user's request will have a corresponding entity. Entity extraction techniques are used to identify and extract different entities. This can be regex extraction, Dictionary extraction, complex pattern-based extraction or statistical extraction. For example, if asked for your favourite colour you would reply "my favourite colour is red". Dictionary extraction would be used to extract the red for the colour entity. What that means in the real world is types of product, locations, model numbers, parts numbers, courses etc: basically anything related to your business which needs to be understood and extracted from the conversation.
-
Intelligent Personal Assistants
This term is often used to describe voice-activated assistants which perform tasks for us such as Amazon Alexa, Google Assistant, Siri etc instead of text-based chatbots.
-
Intent
An intent represents a mapping between what a user says and what action should be taken by your chatbot. A good rule of thumb is to have An intent is often named after the action completed for example FindProductInformation, ReportHardWareProblem or FundraisingEnquiry.
-
Machine Learning
Machine Learning or ML for short is probably used by you every day in Google search for example or Facebooks image recognition. ML allows software packages to be more accurate in predicting an outcome without being explicitly programmed. Machine learning algorithms take input data and use statistical analysis to predict an outcome within a given range. Machine learning methods include pattern recognition, natural language processing and data mining.
-
Natural Language Processing
Natural language processing (NLP) is broadly defined as the automatic manipulation of natural language, like speech and text, by software. NLP is a branch of artificial intelligence that helps computers understand, interpret and manipulate human language. NLP draws from many disciplines, including computer science and computational linguistics to fill the gap between human communication and computer understanding.
-
Natural Language Understanding
A subfield of NLP called natural language understanding (NLU) has begun to rise in popularity because of its potential in cognitive and AI applications. NLU goes beyond the structural understanding of language to interpret intent, resolve context and word ambiguity, and even generate well-formed human language on its own.
NLU algorithms tackle the extremely complex problem of semantic interpretation. That is understanding the intended meaning of spoken or written language. Advances in NLU are enabling us create more natural conversations.
-
Sentiment Analysis.
Sentiment Analysis is the process of determining whether a piece of writing is positive, negative or neutral. More advanced analysis would look at emotional states such as "angry", "sad", and "happy".
-
Utterance
An utterance is anything the user says via text or speech. For example, if a user types “what is my favourite colour", the entire sentence is the utterance.
-
Conversational IVR
Conversational IVR is a software system which uses voice commands from customers. This allows them to interact with IVR systems over telephony channels.
Whereas traditional IVR systems had speech recognition technology to handle simple voice commands such as “yes" or “no". Conversational IVR allows people to communicate their inquiries in more complete phrases via a natural language understanding. Callers can describe questions or concerns in their own words which is then matched to an intent by natural language understanding.
We hope you have found this Conversational AI Terminology Cheat-sheet helpful.
Comment if you think I've missed any terms out which should be on the cheat sheet.
If you want to talk about your chatbot project why not book a free consultation with us.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Dialogflow Migration V2 Guide [Part 1]
***UPDATE***
Dialogflow have extended the V1 API shutdown deadline to March 31st, 2020. https://cloud.google.com/dialogflow/docs/release-notes#November_14_2019
Winter is coming! (for any Game of Thrones fans this will make perfect sense!) In October last year, we wrote about the news that Google will be dropping support for V1 of the Dialogflow REST API in Oct 2019. We've been building all our chatbots with V2 since last year, however, there are many companies who still have V1 Dialogflow agents which will need to be transferred. This blog post aims to help you with carrying out your migration successfully.
The amount of work needed will really depend on what features your Dialogflow agent is using and where it's integrated: If you are using Dialogflow's fulfillment webhook, inline editor, or any Dialogflow API, you'll need to update your code, endpoints, and/or fulfillment to be compatible with V2. However If you are certain your existing agent doesn't use the fulfillment webhook library, the Dialogflow API, or any integrations, then you will not need to make any major changes before selecting V2.
Due to authentication changes, the biggest impact will be for Dialogflow web agent implementations which are currently calling the REST API.
This post will be split this 2 sections: a basic migration guide for agents not using the REST API and a more advanced version covering what changes are needed to use the new REST API and what changes need to be made to support authentication.
You can see more details about upgrading from V1 to V2 in the official guide here.
Basic Migration
Anyone who already has built out their website chatbots using v1 API, then they should start planning for the migration sooner rather than later. Any new features should be added after the upgrade. The migration is potentially a non-trivial task, considering some chatbots have some fairly complex code driving their fulfilment. If you have a live bot in production our advice is to set up an upgrade chatbot as a copy of your existing bot project and then work through the upgrade there. You can guarantee that changing to V2 will mean that fulfilment and API calls may stop working. Once the upgrade is complete re-testing all bot functionality is strongly advised before setting live.
Chatbot Web Interfaces
We would recommend everyone who is creating custom website chatbots to do so using the v2 API. All our new chatbots are built using the v2API.
If you need assistance or advice with your own chatbot v2 upgrade please get in touch, we are Dialogflow experts and would be happy to help!
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Buckinghamshire Business Festival Sponsor 2021
We are proud to be a Buckinghamshire Business Festival sponsor this year
We are proud to be sponsoring the 2021 Buckinghamshire Business Festival, running from April 19th – 30th.
The festival has been organised by Buckinghamshire Business First, with a packed schedule of events and opportunities to make new connections across the two weeks.
Look out for our article in the Buckinghamshire Business First Sponsor Newsletter due to be sent on 8th April – we are excited to be inviting businesses to contact us to find out more about what we do.
Visit the Buckinghamshire Business Festival webpage to find out more about the Festival and to see the full schedule of events: https://bbf.uk.com/events/buckinghamshire-business-festival
Use the hashtag #BucksBizFest on social media to get more involved with the festival in the build-up and as it happens.
Contact Buckinghamshire Business First for more information on the festival: 01494 927130 / events@bbf.uk.com
Sign up for a free conversational AI strategy consultation
This event is part of the 2021 Buckinghamshire Business Festival, running from April 19th – 30th. The festival has been organised by Buckinghamshire Business First, with a packed schedule of events across the two weeks.
“With our 30 minute conversational AI strategy consultation, we aim to identify how conversational AI technology can help your organisation.
Firstly we will start by gaining some understanding of your core business. Then discuss any business or process problems or challenges which can be addressed with a chatbot or voice assistant.
We will tailor our consultation to cover the areas of conversational AI technology and benefits you are interested in and where and how they could be applied to your organisation."
Adrian Thompson
In the meantime you can read learn more about chatbots and voice technology in our blog.
Visit the Buckinghamshire Business Festival webpage for the full schedule of events: https://bbf.uk.com/events/buckinghamshire-business-festival
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
What Makes a Successful Chatbot Project?
Tips. Insight. Offers. Are You In?
Your Chatbot Project
In this article, we will show how we can carry out your successful chatbot project at The Bot forge. We’ll share a few things we’ve learned from building chatbots. We’ll look at messaging platforms, voice interfaces and the importance of conversation design. Giving you a walk-through of how your chatbot project would progress, forging the perfect chatbot.
What should my chatbot do?
Right from the beginning of the project, It's important to have a clear understanding of what your chatbot will do. We like to ask the question:
What is the number one reason my chatbot will exist?
Chatbot platform
Your audience should drive your chatbot platform choice if possible. If you can collect information on which messenger platforms your audience use then this should assist your decision. Facebook Messenger is the most popular with over 1 billion active users as well as being constantly improved by the Facebook team. This is our favourite messaging platform. We particularly like the UI elements which we can provide using the Facebook platform. Have a look at our Customer Support chatbot to see an example of these elements.
If you want your chatbot project to live in a voice interface we recommend either Amazon Alexa or Google Home. The Amazon Echo has brought the voice interface to over 20 million homes around the world. Alexa is not just for handling home tasks Alexa for business means she can help you at work, acting as an intelligent assistant and integrating with your enterprise systems.
Chatbot Requirements
We like to capture chatbot requirements as user stories. The story is in the same format:
As a , I want , so that for example:
- I'm a hotel guest, I want to book a table in the restaurant, so that I can eat tonight
- As a hotel owner, I want to collect a guest reviews, so that I can improve their experience.
- As a previous customer, I want to have similar clothing choices recommended so I can match my style.
Chatbot Persona/Personality
Decide on bot’s personality and tone of voice. This character can then be used in planning the conversation in later steps forming the bot's persona in conversing with users.
We recommend creating a complete agent character which we can then model and grow, possibly adding humor to responses. Our visual designers can assist in coming up with a chatbot character.
Chatbot Conversation Design
We prioritise the user stories and plan them our in more detail. Elaborating as we go through each one. and noting main entry points to back-end system integration.
Design the conversational workflow at a high level. Record all the possible topics and conversation parts, a whiteboard session is great at this point and a simple mind map.
The next step is the conversational ui scripts(cui). We will write the bot scripts incrementally, starting with the core functionality and then expanding into personality driven intents and multiple responses. Always trying to focus on one conversational part at a time. Some examples of the scripts:
Hi there, I'm the macbot ready to give you weather forecasts hi macbot where do you live, so I can send you weather forecasts. leeds ah ok, nice city right now its 1°C I hope you've got a hat and gloves!
As part of the script design we will also elaborate on other elements of the dialogue, for example sets of options, conditions, user input and entities from the input.
We will also define any custom entities for the chatbot. The entities are used for extracting parameter values from natural language inputs. For example the following entities:
- Total Spa Experience
- Experience Wellbeing Massage
- Experience Body Pumice
Then if you said "can I book a total spa experience" the chatbot would be able to pick out the total spa experience in your reply.
{Hi, hello, good-day} I'm the spaworld bot which spa treatment did you want to book? can I book the [body pumice experience] please sure, what time and date? [midday] on the [23rd of feb] ok sure, same as last time right, with Anton? actually is [Denise] available yes no problem, all booked
We will agree on the core conversation dialogues and fine tune them. Then we will also plan on how to handle users straying away from what we call the happy path (following the normal conversational flow). We will also allow for users trying to challenge the bot with sexts, swearing, off context questions, swearing or gibberish. We will always keep an eye on the conversation goals and insure they offer the best user experience whilst making sure the intents match core functionality.
Chatbot UI Design
The BOT Forge will produce prototypes for each core intent as an interactive mock-up providing the visual text interface and voice interactions to show how the conversation will flow.
We also decide on whether we will use different types of structured messages (images, buttons, quick replies, lists, web-views. Depending on the chosen chatbot integration the mock-ups will also include some UI elements specific to the platform. We can share these mockups online or provide them as an animated gif. This is a great way of showing how a bot will work in real chatbot environment, or how the conversation will sound in the voice interface.
Chatbot development
We use an Agile development process using sprints, releasing features little and often to meet the story features. We work closely with our clients, always testing, improving the bot flow, the conversational knowledge base, the bots personality and the overall user experience. This process of iterative delivery a working chatbot will be deployed and ready to use by real users, right from the very first sprint.
Chatbot Platform Integration
We like to carry out the integration work for the chatbot as one of the first pieces of development work. This is so we can provide a working bot for our clients to be able to see a beta/testing version of the chatbot as early as possible. Then we can release bot features and conversation intents regularly.
Conversation Development
The conversational ability will then be implemented in the chatbot following the conversation design and initially focusing on the core intents and text responses. This will be an iterative process. Matching user intents to core functionality and features and training the natural language processor to understand users and handle conversation failure scenarios gracefully. As this stage some integration features will be mocked to return dummy data. During this development stage the chatbot will be provided as a beta implementation so that its available for its first conversations with our client. Users can be notified of new intents for testing.
The training data at this stage will be invaluable for perfecting the bot conversation. This process will also highlight any need for new responses as a continuous cycle. Testing the bot based on responses will continue, we call this supervised learning .
Further UI elements will also be created dependent on the chosen integration platform for the chatbot.
Integration Development
Once the conversational ability has been implemented we will implement any integrations needed for the core chatbot functionality. Writing the code to connect and extend your backend services and integrating with external services needed for the bot to deliver the correct functionality.
Working through each story element and replacing any mocked data entry points. Each piece will be unit and system tested.
Chatbot Alpha/Beta Testing
Once development has been completed we will define how long we are going to do Alpha and Beta testing. Alpha testing is a type of acceptance testing; performed to identify all possible issues/bugs and continued supervised learning , before releasing the chatbot to your users. Bugs will be logged and tracked on our tracking tool and prioritized and fixed on a regular basis. The aim is to carry out the interactions with the bot that a typical user might perform. Making sure to carry out each user story to get the expected outcome. We are happy for our clients to become involved although to be fair you will already of had exposure to the bot as part of the ongoing agile project.
Beta Testing of a chatbot is performed by "real users" of the software application in a "real environment".
Ideally the chatbot Beta version is released to a limited number of end-users of the product to obtain feedback on the product quality. Beta testing reduces conversation and integration related failure risks and provides increased quality of the user experience through customer validation.
It is the final test before shipping the chatbot to your customers. Direct feedback from customers is a major advantage of Beta Testing. This testing helps to tests the bot in real time environment.
The experiences of the early users are passed on to the developers, who make final changes before releasing the bot commercially.
Chatbot Deployment
Once the round of beta/alpha testing has been completed we can deploy the chatbot as a live application.
Chatbot Future
Of course it doesn't end there. Once deployed the chatbot project will be maintained by The Bot Forge as a yearly subscription. We will constantly monitor your bot carrying out daily supervised learning and weekly improvements. Monitoring conversations and confirming qualified intents as well as checking for unmatched intents and fixing them as needed.
We will carry out third party and integration maintenance, monitor api's for version updates. Making sure your bot is performing well and healthy!
The Bot Forge are always available to discuss further improvements and functionality to add to your chatbot or just to talk to us about your next great idea.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
10 Questions To Ask When Planning a Chatbot Project
Tips. Insight. Offers. Are You In?
Planning The Best Chatbot
Congratulations you've had that lightbulb moment and you have an idea to create the best chatbot, or maybe you've heard so much about chatbots lately you feel you should explore the idea of using one for your company. If you want to read more about how a chatbot can help your company read our Why using chatbots for business can help you remain competitive blog post
Whatever the reason now it's time to start looking at your idea in more detail to plan the best chatbot..but wait that can be hard. Don't worry, we can help!
Chatbots have progressed rapidly over the past couple of years with advancements in Natural Language Processing (NLP) utilized used across voice and text-driven interfaces. There has never been a better time to start a conversational UI project. However its still vitally important to plan your project carefully.
At The Bot Forge, we like to ask our clients the following questions to ensure we have a clear understanding of what they want to achieve with their chatbot project.
So whether you are looking at creating your own bot, or commissioning a team of chatbot experts like ourselves ;) then it's important to ask yourself the following 10 questions before you start building the best chatbot.
1. What is the purpose of your chatbot?
Why do you want to create a chatbot? What do you want the chatbot to do for your business and how will it achieve your business goals?
Right from the beginning of the project, it? s important for yourself and your team to have a clear understanding of what your chatbot will do
2. What are the key goals of your bot project?
What are the main aims of your conversational ai project? It could be to drive sales, provide 24/7 customer support or engage with new and existing customers by gathering customer feedback and delivering new product information.
3. How will you measure your success
How will you determine the success of your chatbot? What will your Key Performance Indicators (KPI) be?
For example, you could look at click-through rates, the numbers of inquiries handled correctly or feedback statistics gathered.
4. Who is going to use your chatbot?
Have a clear idea about who is going to use your chatbot, what will be the user demographic? This may influence your chatbot's persona.
5. Where will your chatbot be deployed?
Your audience should drive your chatbot platform choice if possible. You can deploy conversational ai assistants to a lot of places:
- Facebook Messenger Chatbot
- WhatsApp Messenger Chatbot
- Telegram Chatbot
- IBM Watson Chatbot
- Slack Chatbot
- Twilio
- Microsoft Teams Chatbot
- Custom Website Chatbot
- IoT
It really all depends on your use case. If it's an internal tool for your HR team then Microsoft Teams makes sense. Or if you need to help your website users then you can create a website chatbot.
If you can collect information on which platforms your audience use then this can assist your decision. Now is also a good time to consider voice platforms such as Alexa or Google Home.
6. What will your chatbot do?
Here you can really start to consider what sort of functionality the chatbot needs to provide and most importantly the conversations it will be able to support.
A good way to capture chatbot requirements is by looking at them as user stories. The story is in the same format: As a , I want , so that: for example:
- I'm a participant, I want to check what time I can start my event, so that I can be ready to leave in good time.
- As a business, I want to collect customer reviews, so that I can improve their experience.
- As a customer, I want to access my account details quickly and receive an account update through my personal assistant.
7. Will the chatbot have a character?
Will the chatbot have its own persona, will it have a character?
Is the chatbot going to just be a polite assistant or does it need a character to carry through your brand?
8. How will the chatbot create value?
Think about the overall user experience. How will the chatbot ensure that users come back?
For example by providing a simple and well-executed personal assistant then customers are going to use this as their first port of call to find information and/or contact your company.
9. How will people find the chatbot?
How are you going to drive people to find and use your chatbot?
Links on your website and also advertising on Facebook can be great places to start as well as content on your Facebook page.
10. How will you look after your chatbot?
How will you monitor performance after launch? In comparison to other projects, it's important to note that once the chatbot is launched this is just the start of your journey. Essentially you are at the start of the optimization phase. You will need to provide resources to get the most out of your automated assistant after it's gone live.
You will need to monitor user interactions, reactions, unanswered requests: so you can train and improve the overall user experience, training your chatbot is key!
Conclusion
After working through these 10 questions you should be well on your way to understanding your chatbot concept.
With all our new clients at The Bot Forge chatbot agency, we ask them to fill out our chatbot checklist, feel free to download and work through it with other members of your team.
We hope you find this post helpful in getting to grips with your chatbot project, feel free to share if you find the 10 questions useful.
At The Bot Forge, we specialize in conversational AI so why not book a free consultation with us.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
6 Tips to Ensure Your Chatbot is GDPR Compliant
Tips. Insight. Offers. Are You In?
General Data Protection Regulation (GDPR) entered into force and was fully operational as of May 25th 2018. You can read all about it here. The new regulations brought a series of changes and improvements while strengthening the current regulatory framework. The GDPR applies to any website or mobile application collecting data from EU residents and that means chatbots and voice assistants as well!
Despite some myths and misunderstandings around GDPR the regulations there has been some success in the new policy despite still being described as being in a transition period. With incidents such as the Cambridge Analytica scandal last year users are even more concerned as to what we do with their data.
It's important to note that, 71% of UK adults want tougher action in penalising companies that abuse our data privacy by misusing third-party data.
If you use chatbots as part of your sales and marketing strategies, you’ll need to make sure the processes you use to collect consumers’ personal data, as well as what you do with this data are in line with GDPR. Read on for some tips on how to ensure that your chatbots are GDPR compliant.
1. User Consent
Consent is not valid unless it is “freely given, specific, informed, and unambiguous." Basically, that means a “clicked" agreement is required.
For websites, your privacy notice is a great place to get consent from users. Here is a great example:
Don’t forget to update your privacy policy!
One of the rules of the GDPR is that all companies utilizing consumer data need to have a clearly stated privacy policy which contains the following pertinent information:
- What information is collected?
- Who is collecting it?
- Why is it being collected?
- How long will it be used for?
- Who will it be shared with?
- How can consumers withdraw from the agreement to give their data?
For a chatbot, it should provide users with a clear-cut, transparent, distinguishable, and easily accessible form to understand what data is collected, and how it will be used by the bot and organization. This needs to be provided at the start of the conversation and also its often a good idea to provide an easy way to access this in future e.g for bots supporting NLP a free text intent or part of an integration menu such as Facebook Messengers:
We've found that having a privacy page in place listing all the important information is also an effective way to aid in compliance.
2. Allow users to have their data forgotten
According to the GDPR, users should be able to request that all their Personal Data is removed.
Chatbots need an intent to support this e.g ‘please forget my data’, ‘delete my personal data’, etc. Or this could be part of the menu system:
This data removal request needs to be followed up correctly.
3. Allow users to retrieve their data
Users should be able to retrieve their Personal Data.
Chatbot users should be provided with a clear and simple way to access, review and download copies of their data (in an electronic form) that was collected, free of charge. This can be actioned in multiple ways. You could either build a dialogue for this e.g ‘please tell me what data you are storing’, ‘can you send me my data’. The response should present the data to the user or send an email to start the process.
4. Use personal data for the stated purposes only
This is vital for becoming GDPR compliant. Your online chatbot may be an informal way of collecting personal data, but it is still considered to be a data collecting and processing tool and so will fall under the GDPR legislation.
Clearly stating what information is used for is key. This means that you are only able to use the data for the stated purposes, such as sending newsletters, emails, SMS marketing messages or contacting users on Facebook Messenger.
Implement a mechanism to make sure users are clear as to what you will do with their data. This can be added as part of a welcome or supported by intent match or part of the privacy policy.
If you tell your customers that you will be using their email address and mobile phone number to send them information about your services and products, you should do that and nothing more.
5. Leverage Chatbot Conversation
Chatbots provide an engaging interaction medium for users which is no doubt enhanced by a personalised experience. This will often mean that a chatbot needs to collect some personal data from their users. When designing chatbots always remember to keep privacy first in mind. With a chatbot, it is easy to ask for a users permission and explain why you need it because you are already in a dialogue with your user.
Use opportunities when available to clarify and advise users during the conversation.
6. Safeguarding Data
Roles
There are two important roles defined in the GDPR that affect you as a company and the chatbot you build. Firstly, the data controller and secondly, the data processor:
- Data Controller represents the entity which determines the purposes and means of the processing of personal data
- Data Processor represents the entity which processes personal data on behalf of the controller
Data controllers are the decision makers about which personal data gets collected, stored and processed - so most companies are considered controllers!
Chatbots are all about data. If you want to create a solid conversational experience, you need to use Natural Language Understanding (NLU) and dialogue systems. The underlying machine learning algorithms need training data in order to improve and learn. Collecting this data is necessary to train the models and the more data you have the better the bot performs.
Data is essential - but it's also vital to reduce the risk of data breaches and adhere to the GDPR data processing principles.
With GDPR you are prohibited to store this data without explicit consent from users or if there is no legitimate reason to store this data. If you do have a need to store this data to improve your chatbot’s interaction with consumers, you may not do so unless you have explicit consent.
It’s common for many web and messenger servers to keep different types of logs, such as access, error or security audit logs. These logs might hold personal data such as IDs, IPs, and even names.
Reviewing your logs will allow you to find any personal data and deal with it accordingly.
Cloud Compliance
At The Bot Forge we use the Dialogflow natural language processing engine to create our chatbots. Using Google Cloud services means we can rely on GDPR being upheld with regards to our chatbot data:
At Google Cloud, we champion initiatives that prioritize and improve the security and privacy of user data. We’ve made multiple updates to ensure that Google Cloud customers can confidently use our services now that the GDPR is in effect.
We have peace of mind as compliance with the GDPR is a top priority for Google Cloud. It's important to have this confidence when using third-party services which handle your data.
Want to talk about GDPR and data privacy?
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Dialogflow Knowledge Connectors
Tips. Insight. Offers. Are You In?
In this post, I'm going to look at the new Knowledge Connectors feature in Google Dialogflow. As I look at the features in more detail I'm assuming you understand the more common Dialogflow terms and features - agents, intents & entities.
It's also important to remember this feature is in beta.
The Problem
We've been working on chatbot projects for many years now and a large number of our chatbot projects have shared a similar requirement: the ability to answer a large number of questions on a particular subject. This may be to answer technical questions about a product offering or to offer information for a particular service.
Often the information related to these types of questions is held on our chatbot customer's own websites as FAQ pages or in specific PDFs or unstructured text documents.
These types of knowledge bases can often hold large amounts of information and so technically they will provide answers to thousands of chatbot questions.
The challenge for a successful chatbot is utilising this often unstructured information to understand a question and provide the correct answer. To meet this challenge we can look at 2 approaches; the traditional one and using the new Dialogflow Knowledge Connectors.
The Solution
Traditional Approach
Stepping back a bit it's important to briefly go over the traditional approach to creating chatbot conversational ability. There are a number of different chatbot frameworks out there such as Google Dialogflow, IBM Watson, Microsoft Bot, Rasa etc and they all largely use the same concepts. A user submits a voice or text query and this utterance will be matched to an intent and any entities extracted. The matched intent would either provide a static response or rely on some form of application layer to perform the required action to provide the response to the user.
This approach can be easy. However, things can get complex and difficult to manage if the scope of intents is very large and or/ the information is constantly being updated. If we want to support questions with knowledge base information then each question needs to be created as an intent and the correct response formulated. This can lead to problems such as:
- Problems with the Intent Classification model grow causing more incorrect classifications.
- The amount of effort required to keep adding more training data to the model to ensure that the accuracy of the Intent classification remains high. Fortunately, Dialogflow provides a training UI in the web console to help keep track of any misclassified utterances, analyzing them and adding these to the training data, however, this does take time.
- Creating and managing intents to support new information in document stores.
Knowledge Connectors
Knowledge connectors are a beta feature released in 2019 to complement the traditional intent approach. When your agent doesn't match an incoming user query to an intent then you can configure your agent to look at the knowledge base(s) for a response.
The knowledge datasource(s) can be a document(currently supported content types are text/csv, text/html, application/pdf, plain text) or a web URL which has been provided to the Dialogflow agent.
Using Knowledge Connectors
To be able to use knowledge connectors, you will need to click "Enable beta features and APIs" on your agent's settings page.
Its also worth mentioning that Knowledge connector settings are not currently included when exporting, importing, or restoring agents. I'm hoping that this is something currently being put in place by the Dialogflow team.
Knowledge connectors can be configured for your agent either through the web console or using the client library that is available in Java, node.js & python. You can also configure from the command line.
To create a knowledge base from the web console, login to Dialogflow & then go to the knowledge tab. The process is fairly straightforward and involves providing a knowledge base name then adding a document to the knowledge base.
You can read more information creating a knowledge base here
After you've done that then you just need to add an intent and return the response. It's also worth keeping in mind you can send all the usual response types and that means including rich responses which I think is pretty cool.
Trying out knowledge connectors
Ok, so its time to try out these wondrous new knowledge connectors. There are 2 different types of knowledge base document type: FAQ & Extractive Question Answering. These choices govern what type of supported content can be used. There are also a number of caveats for each content type which you can read more about this here
Based on these 2 document types I looked at a couple of common use cases which we often encounter at The Bot Forge and correlate well with the document types supported:
- Chatbot FAQ functionality using an existing FAQ webpage in a fairly structured format to provide answers from.
- Chatbot FAQ functionality using information in an unstructured format to provide answers from.
I carried out my tests using a blank Dialogflow agent with beta features enabled.
1- An FAQ Knowledge Base (Knowledge Type: FAQ)
For my knowledge base I used the UCAS Frequently asked questions webpage and used the following URL as my data source. This processes the URL which is in the correct format and creates a series of Question/Answer pairs which can be enabled or disabled in the console, pretty neat!
So giving this a spin my first test was "how do I apply" and the result was spot on,
matchConfidenceLevel: HIGH matchConfidence: 0.97326803
Whilst different variations on the same question also returned a good result.
"im not sure how to apply" matchConfidenceLevel: HIGH matchConfidence: 0.9685159 "can you tell me about how I can apply" matchConfidenceLevel: HIGH matchConfidence: 0.968346
Unfortunately, when I try something a bit less obvious. I get an incorrect result as it matches the wrong intent.
"how do I submit my application" matchConfidenceLevel: HIGH, matchConfidence: 0.9626459
In this case, it's matching the "How can I make a change to my application" intent with a high confidence but unfortunately it's the wrong intent. So the problem here is we need to fine-tune the model and re-assign the training phrase (utterance) to the intended intent. The limitation is that in the knowledge base you can't fine-tune responses. If you want more control you will need to move this faq over to its own intent.
This problem is compounded by the fact that the training feature of the console just lists each response intent as "Default Fallback Intent". It's hard to check which responses have been answered incorrectly. One way round is to look in the History area of the console and look at the Raw interaction log of each response.
One really useful feature is that you can assign a specific extracted FAQ from the knowledge document and assign to an intent. Just click on view detail in the document list -> select the question and click the "convert to intents button". At the same time, this will create a new intent and disable the current Question/Answer pair. So overall pretty impressive if you have webpage or doc of structured FAQs you can use this to power an FAQ chatbot pretty effectively with some monitoring.
2-A more unstructured FAQ Knowledge Base (Knowledge Type: Extractive Question Answering)
In this use case, I wanted to try out the ability of the knowledge connectors to return answers from more unstructured data.
Again there are caveats about what data source you can use you can read more about this here.
For my test, I used a standard drug leaflet with MIME type PDF covering Priorix, from www.medicines.org.uk. I created a new knowledge base, added a new document and made sure I selected the knowledge type as "Extractive Question Answering". Once imported the PDF is listed in the document list. My aim was to validate if Dialogflow could extract some fairly simple answers from the document. Now for some testing:
"What is Priorix" matchConfidenceLevel: HIGH matchConfidence": 0.88257504 answer : "Priorix, powder and solvent for solution for injection in a pre-filled syringe Measles, Mumps and Rubella vaccine (live)"
Unfortunately, although the response had a high confidence and match score it was actually an incorrect response. Ideally, the answer should have been:
"Priorix is a vaccine for use in children from 9 months up, adolescents and adults to protect them against illnesses caused by measles, mumps and rubella viruses."
I tried another test:
"how is priorix given" matchConfidenceLevel: HIGH, matchConfidence: 0.8826 answer: The other ingredients are: Powder: amino acids, lactose (anhydrous), mannitol, sorbitol
Again this was an incorrect response. I would have expected the correct response to be:
"How Priorix is given
Priorix is injected under the skin or into the muscle, either in the upper arm or in the outer thigh."
So unfortunately not great results in extracting answers from the PDF I used. It would be interesting to look at a selection of other types of documents and corpora.
Do Knowledge Connectors work?
Again it's important to point out this is a beta feature. There are definitely challenges and in some functional areas much more to be done with Knowledge Connects. In conclusion, It's also important to recognise that I looked at 2 different types of use cases and knowledgebase document types which provided very different results so its worth looking at each one separately.
Chatbot FAQ functionality using an existing FAQ webpage in a fairly structured format.
If you want to convert your FAQ page into a chatbot or if you have a similar structured document such as a PRFAQ for a product or service then Connectors work well.
Just supplying the URL of the FAQ page as a data source to the knowledge connectors is fantastic and provides fairly good results. However, it's worth keeping in mind there may still be match errors so the history log is invaluable in checking for them. Thankfully it's fairly easy to manage any question/answer pair which has been handled incorrectly by converting to its own intent.
Chatbot FAQ using a document in an unstructured format.
I found my test results with this use case rather disappointing. The accuracy of the extracted answers was fairly poor for my test case. Although for different document sources you may be able to get better results.
The extracted answers look more like a match based on keywords with some additional coverage but it does not appear to consider the context in which the question is asked. Also, this type of knowledge connector does not provide any full control like intents in terms of context and priority of matching training phrases etc so there is no way of fixing bad responses. A feature where you can evaluate and train responses would be a great addition to the knowledge base so hopefully, that is in the Dialogflow team pipeline.
Should I use Dialogflow Knowledge Connectors?
If you have some FAQ information in a structured format then Knowledge connectors are worth a try with some caveats.
If you have unstructured documents which you want your chatbot to use to extract answers to questions then at the moment knowledge connectors are not a magic bullet. It's a big ask, but for me, this is where the real value will lie particularly if you want to support large knowledge bases with a chatbot. Knowledge connectors are an experimental feature, so hopefully as the technology advances then they will improve.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
How To Create The Perfect Google Business Welcome Message
Why Google Business Messages?
Google Maps has 155 million monthly users and it's estimated that Google handles 5.6 billion searches per day - two trillion searches a year!
Connecting with your customers at these two touchpoints is more beneficial than ever.
Gone are the days when you needed to send customers to a website or social media profile to contact you. Google Business Messages allow you to chat to customers directly in a Google search result (also known as a SERP, or Search Engine Results Page), on Google maps, or through your Google Business Profile (formerly known as 'Google My Business', or 'GMB').
With automation and chatbots for Business Messages in their relative infancy, we've included some great example welcome messages, and tips to help you get the most from your Google's Business Messages chatbot integration.
5 Tips To Help Create The Perfect Welcome Message
Improving user onboarding experiences and conversion rates of your users has never been as crucial as it is now.
A welcome message could well be the first contact a customer will have with you.
The two are unavoidably linked!
Writing better welcome messages for your chatbot solutions can really help set the tone for your customers, and establish a great experience.
These tips will help you to write more effective welcome messages for all your chatbots!
1. Greet, Introduce & Encourage
The major aims of your Google My Business chatbot welcome message should be to greet your new user, explain your chatbot's purpose, and encourage them to take action.
Use your welcome to set expectations for your customers and guide them through the initial stages of their interactions with your company.
2. Keep it Simple
Your message should only be a maximum of a few lines.
Avoid large chunks of text - try and split your content down, naturally, into separate messages.
3. Define Your Chatbot's Persona
Always make sure your users know they are interacting with a chatbot.
Humanizing your chatbot makes the whole experience more usable and fluid - if you don't convey a persona, your users will decide on it for you.
If you don't convey a persona, your users will decide on it for you
Your welcome message is the first opportunity to introduce your chatbot's personality. The language you choose sets your assistant's tone and values - which should match those of your brand.
PS - now's a good time to give your assistant a name. Keep its name short, memorable, and easy to spell & pronounce.
4. Use Emojis (But Not Too Many)
A great way to humanize your Google Business Messages chatbot is to incorporate emojis into your conversation.
Not only are emojis a great way to add space to your text, but they can also add small personality traits to match the tone of your message - whether that be amicable or exciting!
Be sure to stay true to your chatbot's tone of voice, and think about how appropriate they are in context. Emoji are difficult to use effectively in corporate or stressful/sensitive settings, e.g. funeral directors, emergency dentists, or divorce lawyers.
Whatever you go for, go easy. Too many and you'll end up looking spammy - stick to the occasional one here and there to add a visual break, rather than something to focus on.
5. Make Your Call To Action (CTA) Clear
In general, users won't have a lot of time to explore your chatbot's capabilities.
It's wise to make your call to action as clear as possible - you simply can't afford for customers to be confused when they are presented with your chatbot's welcome message.
When designing your call to action, keep your key business and marketing goals in mind.
Making use of Google Business Message's conversation starters is a good idea to follow up with, so users can easily select your CTAs.
You can use up to 5 of these, e.g. "Choose your bike" | "View bikes" | "See latest offers"
Tips. Insight. Offers. Are You In?
5 Awesome Welcome Message Examples
Now we've looked at some tips to consider, let's dive into our 5 awesome examples.
1. Conversational Commerce
This welcome is aimed at increasing sales for a bike manufacturer:
"Hi I'm {Bot’s name}, the GoodBikes virtual assistant.
I can help you, "Choose the correct bike for your riding", "View our bikes", "View latest bike offers".
How about highlighting an offer to drive sales:
"Hello, {Customer’s first name}! I’m {Bot’s name}, and I’d be happy to help you win a 25% discount on your first purchase with {Your brand’s name}. Can I help you find a particular trainer?"
"Road | Trail | Track"
2. Announcements
Offer current data that enables self-service for users.
Change the welcome message and conversation starters to temporarily contain this information if there is a big development or event that you anticipate people would look for, like a service outage.
"Hi, I'm the Livewire virtual service agent"
" I've got one thing you need to be aware of, we currently have an outage; normal service for your area is due to be resumed at 5:00 PM"
"Is there anything else I can help you with today?
3. Contextual Welcome Messages
Contextual information can be used to personalise your greeting.
The user's name, location, entrance point, and place ID are all included in your contextual data (for location-specific entry points).
For each language and place of business that you support, you're able to design a special welcome message:
"Hi I'm {Bot’s name}, the {your brands name} virtual assistant for {location}. I can answer questions about {location}, how can I help?
Closing time | Services | Parking Info"
4. Customer Support
Provide an extra level of support for your customers and manage expectations with response times.
“Hello, {customer’s first name}! Thanks for your enquiry.
Please expect a response from our support agent within 24 hours. In the meantime, why not take a look at our product tutorial: {link to tutorial}"
5. Event Specific
Your welcome messages can be seasonal or tied to promotional events; use your welcome message to highlight this to the user:
"Welcome to {your restaurant name}, I'm {chatbot name} the virtual assistant. Don't forget its burger night tonight, 20% off all burgers! What can I help you with?
Call us | Order online | Book a table"
Conclusion
Following these tips will help you stand out from the crowd and get the attention of your ideal consumers as well as help your existing ones.
Utilizing Google's Business Messages for your brand is surprisingly easy.
All you need is the help of a Google Business Messaging API partner. Knowledgeable partners, such as The Bot Forge, will help you provide rich conversational messaging solutions that will help you increase client loyalty and engagement
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
3 Books That Will Boost Your Chatbot Knowledge
Introduction
Currently, chatbots are dominating online markets, especially in countries such as the U.S., India, Germany, Brazil, and the UK.
According to a Business Insider article on chatbot statistics, 40% of internet users worldwide prefer chatbots over virtual agents because they get answers quickly and more conveniently due to their 24-hour service.
Due to the increasing demand of consumers to have round-the-clock digital experience, experts predict that retail companies will increase their budget for chatbots to $142 billion in 2024 from just $2.88 billion in 2019.
Therefore, we can expect that chatbots will become the primary communication channel for online consumers in the foreseeable future.
As such, here are three books that can jumpstart and enrich your knowledge in the world of chatbots.
1. Designing Bots: Creating Conversational Experiences
Before starting your journey, it is important to understand and familiarise yourself with the fundamental information about chatbots.
Designing Bots: Creating Conversational Experiences by Amir Shevat can help you on this one.
The book introduces the readers to the description, process, and purpose of chatbots. Shevat emphasises that chatbots are not trivial projects. Therefore, businesses that want to utilise chatbots must learn how to effectively use them to ensure a successful implementation and a return on investment.
Our previous article, Planning the Best Chatbot, concisely breaks down the necessary steps when planning a chatbot project, including identifying the purpose, goals, and performance indicators of your project. Alongside our insights, the book provides information on how to build and design your chatbot by presenting actual cases.
With the help of this book, you can learn and refresh your memory with the basic principles and start delving into the deeper concepts about chatbots without feeling overwhelmed.
2. Algorithms to Live By: The Computer Science of Human Decisions
The second book, Algorithms to Live By: The Computer Science of Human Decisions by Brian Christian and Tom Griffiths, explores the power of precise algorithms in helping us decipher human questions, which allows us to predict human choices using computer science.
The book posits that algorithms aren’t limited to the world of computers, but that they also bridge technology and human interfacing. This is why the book is worth reading.
It highlights how mathematical philosophy and life are not so different when it comes to solving problems and making decisions. By solving human problems like mathematical problems, which includes inputting various information and seeing how they work together, we can reduce and deduce possible options and end up with the right answer– a process paralleling that of chatbots'.
Solving problems and the steps that involve them are algorithms themselves, and chatbots work similarly. After customers input their concerns, the algorithm recognizes the problem and presents the most viable solution based on a predetermined set of actions or rules.
Tips. Insight. Offers. Are You In?
3. Business of Bots: How To Grow Your Company Through Conversation
This book, Business of Bots: How To Grow Your Company Through Conversation by Mariya Yao and Adelyn Zhou, goes beyond the theories of chatbots and discusses how businesses who want to connect with their customers can utilise chatbots.
This is best for those who have gone through the chatbot basics and want to advance their knowledge.
According to the authors, chatbots powered by artificial intelligence will not only help businesses improve their customer service departments, but also boost their sales and marketing strategies.
To visualise the success of chatbots, the book also examined and featured hundreds of actionable bot strategies of leading brands in the marketplace, such as Sephora, Expedia, Victoria's Secret, Capital One, and eBay.
After learning the ropes of chatbots, it is time for you to apply that knowledge and enhance your business with chatbots.
Conclusion
In a world where consumers prioritise efficiency and speed, chatbots will become the new mode of communication with consumers.
Understanding how they work and how to best utilise this technology will definitely help businesses keep up with consumer needs. If you need assistance in initiating your chatbot services, do read about our custom chatbot development services.
Post especially contributed by Cara Ariella Erickson for thebotforge.io
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
The Bot Forge is One of the Most Reviewed UK AI Companies in 2022
The Manifest Recognizes The Bot Forge as One of the Most Reviewed AI Companies in the UK
The Bot Forge creates modern solutions to improve organizational efficiency for our partners. Our team of experts helps you design, build, launch, or support enterprise-grade chatbots, voice assistants, and conversational IVR solutions. We aid you in making the most out of AI technology to enhance your business performance.
Today, we’re thrilled to share that we’re among the leaders on The Manifest this year. According to the B2B research on The Manifest, we’re one of the most reviewed AI companies in the UK.
“We are really excited to have been chosen as one of the leading chatbot and voice assistant agencies in the UK by The Manifest."
— Adrian Thompson, Founder of The Bot Forge
We are industry experts in conversational AI, and we’ve been committed to delivering AI architecture expertise to a global client base since 2018. Over the years, we’ve been successful in producing groundbreaking solutions for many organizations worldwide. This award showcases our unyielding efforts in the past years to provide advantageous technology to our partners.
In August 2022, a marketing analytics consultancy partnered with us for the development of a proof of concept. The client needed a virtual agent built on Dialogflow CX designed to illustrate the potential use case of conversational AI in the digital customer journey. The client shared the following about our partnership:
“I appreciated the creative approach because speed and lean focus are the order of the day for POCs."
— Jonathan Lewis, CEO, Marketing Analytics Consultancy
Thank you to Jonathan for taking the time to write his honest feedback. Don’t forget to browse through The Manifest, a company-listing platform, to discover more about our work.
If you’re interested in our AI solutions, please schedule a free consultation with us today.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
How Much Does it Cost To Build a Chatbot in 2023?
Tips. Insight. Offers. Are You In?
Like most software projects the chatbot cost really depends on the scale and complexity of the project.
These requirements will govern the effort involved in building your perfect automated assistant and the ongoing effort required to keep it running smoothly.
Let’s take a look at some of the governing factors in how much your chatbot project will cost.
How much does it cost to create a chatbot to fix our [insert business problem]?
We are often asked this question by clients looking to start their first conversational AI project.
At this point, we tend to ask a specific set of questions to get an idea about the scope of the project they have in mind. You can read a bit more about your chatbot scope here.
Chatbot cost can be broken down into 2 parts.
- One off design, development, deployment of the chatbot
- Ongoing monthly maintenance, hosting and management costs
Spoiler alert: we can’t give an exact price without knowing the details of your project. However, we can give some estimates based on the type of project. Jump to the end if you'd like to go straight there.
Let’s dive in and take a look at the key drivers impacting the price for custom chatbots.
Deployment Channel Chatbot Cost
The first requirement to consider is the channel in which the bot should function.
By this, I mean where the bot will be used.
It could be as a website widget implemented on your webpage or deployed into existing messaging platforms.
Examples of these platforms include Facebook Messenger, WhatsApp, Microsoft Teams, Slack, Telegram, and Viber. Some of these examples provide more complex UI elements which can be utilised in your bot.
Other obvious channels to deploy a chatbot are voice; Google Assistant, Alexa or even Conversational IVR systems which can be connected to the same conversational engine as your text-based chatbots although voice assistants will invariably need work specific to voice channels.
For our custom chatbot integrations, we normally provide one channel with the project and then charge per extra channel, as required.
We’ve found that our clients will often want a web-based chatbot first and create and deploy different versions to other channels depending on the chatbot use case.
Extra channel development costs vary, depending on the amount of work required to create the best user experience for the platform.
Features
Depending on where the chatbot will work in there is also scope to provide other functionality such as voice capability for web chatbots or enhanced chatbot interface features specific to the deployment platform.
Again, costs depend on the amount of complexity and effort involved in building each feature.
As an example, adding voice interaction capability to a web chatbot would be £3,000+.
Human-agent handover via live chat or WhatsApp integration is also a popular feature.
Natural Language Ability Chatbot Cost
If a chatbot is required to support more complex natural language understanding (NLU- you can read more about some of the tech terms here) features and not just UI elements such as buttons then this will mean that additional effort is needed to train the bot and design and implement a more complex conversational flow.
In reality, most text-based chatbots will have some level of NLU and, of course, voice assistants are all NLU.
We utilise the best of breed NLU solutions to create our conversational experiences.
In particular, we use Google Dialogflow ES and CX for many of our projects and, as a result, we are Google Tech partners and experts in Dialogflow. For onprem or open source solutions we also use Rasa.
Conversation Skills
The complexity, scope and volume of the required conversational ability also affect cost. This relates to specifics such as:
- The number of branches in the conversation tree
- The number of questions that have to be handled by the chatbot, which can often be in the thousands
- The number of training phrases needed; this element can improve the accuracy of the chatbot, (you can read more about training data here)
- The complexity of conversational ability i.e. support for complex user enquiries, multi-turn conversations.
- Number of entities needed to support the required responses, don’t forget you can remind yourself of terminology here
Languages
Chatbots are capable of supporting different languages, as long as these are supported by the NLP engine.
It’s possible to add different language permutations to the same chatbot project. However, each language will need its own testing and all the responses for each language and potentially any responses returned by business logic may need to be altered for each language.
A conversational designer will also need to consider the nuances of each language here, so costs for each language will depend on the size and complexity of the conversational ability for the primary language.
Integration Chatbot Costs
Connect your chatbot to existing systems: APIs, RPA, Knowledge Bases
The other area which will impact cost is dependent on the planned role of the chatbot: what the chatbot will need to do to carry out its role?
Will the chatbot need to integrate with current systems to provide its responses?
Will it need to hand over to live agents?
Will it need to connect with CRM and ticketing solutions?
Some chatbots may need to carry out complex interactions to provide answers to customer queries.
You can read more about possible integrations here, but the rule of thumb is that if the system you want to integrate with has an API and a means of authenticating then we can integrate with it.
Chatbots can also leverage other AI systems to provide relevant information to govern conversational flow. For example, sentiment analysis.
With so many possibilities for chatbot features, it's hard to estimate the price here as each integration can have its own complexities and sets of APIs to authenticate with.
Integration Chatbot Costs
Connect your chatbot to existing systems: APIs, RPA, Knowledge Bases
The other area which will impact cost is dependent on the planned role of the chatbot: what the chatbot will need to do to carry out its role?
Will the chatbot need to integrate with current systems to provide its responses?
Will it need to hand over to live agents?
Will it need to connect with CRM and ticketing solutions?
Some chatbots may need to carry out complex interactions to provide answers to customer queries.
You can read more about possible integrations here, but the rule of thumb is that if the system you want to integrate with has an API and a means of authenticating then we can integrate with it.
Chatbots can also leverage other AI systems to provide relevant information to govern conversational flow. For example, sentiment analysis.
With so many possibilities for chatbot features, it's hard to estimate the price here as each integration can have its own complexities and sets of APIs to authenticate with.
Deployment & Infrastructure Security
Often security demands for a chatbot project need specific features, for example, HIPAA compliance.
In these cases, SSO, RBAC, and on-prem or private cloud deployment can be used to ensure compliance with company security policies.
These can have an impact on overall project costs and again, costs are based on the demands of a specific project.
Chatbot Training & Maintenance
We offer our chatbot solutions based on a SAAS model.
Costs incurred tend to be based on a yearly subscription and again depend a lot on the scale and complexity of the chatbot.
These monthly costs will cover the following:
- Access to our world-class chatbot and voice assistant analytics platform (Chatseer)
- Daily supervised learning and improvements
- Monitoring conversations and confirming qualified intents as well as checking for unmatched intents and fixing them as needed
- Third-party and integration maintenance. Making sure your bot is performing well and healthy!
- Natural Language Understanding service costs (depending on volume and platform used).
- Hosting and data storage
- Chatbot reporting interface
- Post-development support.
As a rule, monthly maintenance costs tend to be in the region of 10% of the initial implementation cost.
Chatbot Packages
The cost of a chatbot project can vary widely depending on the overall scale of the project and the features required.
We tend to split our projects into 3 packages. You can see the features included and the one-off project costs and monthly costs in the table below.
It’s worth keeping in mind that the cost of a capable chatbot does not have to be prohibitive and it’s often easy to start small and add features as business needs require them.
We make sure all our chatbot and voice assistant projects will scale. So even if you want to start with a smaller scale chatbot solution, your company can still expand and build on this to create a large scale solution further down the line.
Discovery and requirements phase
We provide a discovery and requirements service for conversational AI projects. Our aim is to help businesses understand the potential of conversational AI and identify the requirements for a specific project.
This service typically includes an analysis of the business needs and goals, an assessment of the available data and resources, and the identification of potential challenges and opportunities.
We will work closely with you to understand the problem that needs to be solved, and to identify the key requirements for the conversational AI project. We will also recommend different conversational AI platforms and supporting technologies that can be used to build the solution.
Our goal is to help your organisation understand the potential of conversational AI and to provide you with a clear plan for how to proceed with the project. The cost of the service will depend on the complexity of the project and the level of support required from the service provider.
Deliverables
Project Cost: £3,000 to £15,000
Proof of concept (POC)
If you have an idea or use case for a conversational AI product or feature it's normally good practice to create a POC.
A POC is an early model that does not have all the final product's functionality, the main goal of a POC is to test the technical feasibility of a solution, and to identify any potential challenges or issues that would need to be addressed before moving forward and investing time and money on the development of a full-fledged system or application.
We aim to keep costs down but ensure all the work can be used as a basis for a production project.
Deliverables
- Define the problem or use case
- Identify the data and resources required
- Analyse and select conversational AI platforms and supporting technologies
- Build a simple conversational AI model
- Create integration for the chosen channel
- Carry out all required integration works
- Test the POC
- Evaluate the results
- Iterate and improve the model based on the results
Project Cost: £2,500 to £15,000
Small Project
A smaller chatbot project is a relatively simple and straightforward implementation of a conversational AI system, typically designed to address a specific use case or business need. This type of project may involve building a chatbot that can handle simple customer inquiries, such as answering FAQs or providing information about products or services.
The chatbot can be built using pre-built chatbot platform or framework, with pre-trained models that can handle natural language processing (NLP) and understand the user's intent. The chatbot can be integrated with the business's website or mobile app, and can be accessed by customers through a chat interface.
The chatbot's functionality can be limited to the specific use case that it is designed to address, such as providing customer support or information about a particular product or service. The chatbot's responses can be pre-defined, and the chatbot can be trained to understand and respond to a limited set of customer queries.
The cost of a smaller chatbot project will depend on the complexity of the use case, but generally, it is relatively low compared to a larger, more complex conversational AI project. Additionally, a smaller chatbot project will typically require less time and resources to develop and launch, and may be used as a stepping stone to larger, more complex projects in the future.
Project Cost: £2,500 to £10,000
Monthly Maintenance Cost: £200 to £1000 (depending on volumes)
- Website chatbot
- Facebook Messenger chatbot
- NLU
- 10-20 intents
- 100s training phrases
- 5 – 10 rich UI elements
- knowledge base support
- Small talk
- 1 language
Medium Project
A medium chatbot project is a more complex implementation of a conversational AI system that addresses multiple use cases or business needs. This type of project typically involves building a chatbot that can handle a wider range of customer inquiries and provide a more personalized experience.
The chatbot can be built using pre-built chatbot platform or framework, with pre-trained models that can handle natural language understanding and complex conversations. The chatbot can be integrated with multiple channels such as website, mobile app, social media platforms etc.
The chatbot's functionality can include multiple use cases, such as providing customer support, answering frequently asked questions, providing information about products or services, and even handling transactions. The chatbot's responses are generated through a combination of predefined and dynamic responses, and the chatbot can be trained to understand and respond to a wide range of customer queries.
The chatbot can also be integrated with other systems such as CRM, ERP, or inventory management systems to retrieve information and perform actions. The chatbot can also be equipped with features such as personalization, sentiment analysis, and analytics to provide a more engaging and personalized experience for the customers.
The cost and time required for a medium chatbot project will be higher than a smaller chatbot project as it involves more complexity, advanced features and integrations. However, this type of project can bring significant benefits to the business, such as cost savings, improved customer engagement, and increased efficiency.
Project Cost: £10,000 to £25,000
Monthly Cost: £500 to £2,500 (depending on volumes)
- Website chatbot
- Facebook Messenger chatbot
- Microsoft Teams Chatbot
- Slack Chatbot
- Alexa Skill
- Google Assistant
- NLU
- 1000s of training phrases
- 50-100 intents
- 10-25 rich UI elements
- Simple integration
- 1 or 2 languages
Large Project
A large scale or enterprise chatbot project is a comprehensive implementation of a conversational AI system that addresses multiple business needs and use cases across an organization. This type of project typically involves building a chatbot that can handle a wide range of customer inquiries and provide a more personalized experience, and it can integrate with multiple internal systems and processes.
The chatbot can be built using pre-built chatbot platform or framework, include extensive bespoke coding and complex integrations. The chatbot can be integrated with multiple channels such as website, mobile app, social media platforms, SMS, smart Interactive voice response (IVR), and even voice assistants like Alexa, Google Home or even digital humans.
At this level functionality can include multiple use cases, such as providing customer support, answering frequently asked questions, providing information about products or services, handling transactions, and even automating internal and external processes such as smart IVR integrations.
The chatbot can also be integrated with other internal systems such as CRM, ERP, inventory management, and HR systems to retrieve information and perform actions, as well as with external systems like payment gateways, and logistics providers.
Large scale or enterprise chatbot projects are complex and require significant resources and expertise. They can be costly, but they bring significant benefits to the business, such as cost savings, improved customer engagement, increased efficiency, and automation of internal processes. Additionally, a well-designed and executed enterprise chatbot can provide a competitive advantage and can help in differentiating the business from its competitors.
Project Cost: £25,000 to £100,000+
Monthly Cost: £2,500 to £5,000+ (depending on volumes)
- Large scale enterprise chatbot with multiple API integrations
- 10,000s of training phrases
- Personalised UX
- 1000s intents
- Multiple knowledge bases
- Deployed to multiple channels including conversational IVR
- Custom user interface elements
- Bespoke functionality
- Multiple languages
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Arabic NLP Guide [2023 Update]
Introduction
Arabic is the fourth most spoken language on the internet and arguably one of the most difficult languages to create automated conversational experiences for, such as chatbots. An Arabic chatbot is a program that can understand and respond in Arabic.
Natural language technologies enabling us to simulate and process human conversations in Arabic have improved a lot over recent years. Enabling us to train to understand the emotions, and meanings, and detect the misspellings and sentiments of the language.
In this post, we wanted to take a look at the challenges, and available tools and create a brief proof-of-concept chatbot using one of these tools.
Arabic NLP Challenges
Arabic natural language processing (NLP) is a rapidly growing field, but it also presents a number of unique challenges compared to other languages.
- Sparsity of Data: One of the biggest challenges facing Arabic NLP is the lack of large-scale, labeled datasets. This makes it difficult to train accurate models and leads to low performance on certain tasks.
- Complex Script: Arabic script is complex and includes many diacritics and ligatures, which can make text pre-processing and feature extraction more difficult.
- Morphological Complexity: Arabic has a complex morphological structure, which can make it difficult to accurately segment words and identify the root of a word. This can make tasks such as stemming and lemmatization more challenging.
- Language Variation: Arabic is spoken in many countries and dialects, which can lead to variations in vocabulary, grammar, and syntax. This can make it difficult to design models that are able to handle the diversity of the language.
- Annotation Challenges: Annotating text for NLP tasks is always a challenge, but it is even more so for Arabic due to the complexity of the language and the lack of resources.
- Right-to-Left Script: Arabic script is written from right to left, which can make it challenging to integrate with left-to-right script systems and can also affect text alignment and layout.
- Lack of Standardization: There are few standard resources for Arabic NLP, such as corpora, part-of-speech tag sets, and named entity recognition tags, which can make it difficult to compare results across different studies and to replicate previous work.
- Cultural and Religious Sensitivity: Arabic text may contain sensitive cultural and religious topics, which may require special consideration when processing and analyzing the data.
Despite these challenges, there is a lot of ongoing research and development in the field of Arabic NLP, and many organizations and researchers are working to overcome these obstacles. With the increasing demand for Arabic NLP in areas such as customer service, e-commerce, and social media, it is important to continue to invest in this field and develop solutions that can help organizations to better understand and engage with Arabic-speaking customers.
To conclude, Arabic NLP is challenging due to the complexity of Arabic script and grammar, the lack of data, and the diversity of the language.
Arabic Conversational AI Technologies
The NLP technologies include advanced machine learning algorithms, natural language understanding models, and language-specific libraries and tools which need to carry out the following tasks:
- Arabic Speech Recognition: This technology is used to convert spoken Arabic into text, which is then processed by the conversational AI system.
- Arabic Text-to-Speech: This technology is used to convert text-based input into spoken Arabic, allowing the chatbot or voice assistant to speak in the language.
- Arabic Natural Language Processing (NLP): This technology is used to understand and interpret the meaning of text written in Arabic. It includes techniques like tokenization, part-of-speech tagging, and sentiment analysis.
- Arabic Language Modeling: This technology is used to train machine learning models on large amounts of Arabic text, allowing them to understand and generate the language.
- Arabic Sentiment Analysis: This technology is used to determine the emotions and opinions expressed in Arabic text, which is useful for understanding customer feedback or gauging the effectiveness of marketing campaigns.
Technical Solutions
CAMeL Tools
CAMeL Tools is a suite of Arabic natural language processing tools developed by the CAMeL Lab at New York University Abu Dhabi.
The camel-tools package comes with a nifty ‘morphological analyzer’ which — in a nutshell — compares any word you give it to a morphological database (it comes with one built-in) and outputs a complete analysis of the possible forms and meanings of the word,
The tool will reduce orthographic ambiguity to account for several common spelling inconsistencies across dialects. Camel-tools accomplishes this by removing specific symbols from specific letters.
Repustate
The Repustate platform provides a number of natural language processing tools for analyzing Arabic dialects. It understands three major Arabic dialects – Gulf Peninsular, Egyptian, and Levantine Arabic also it Obtains granular Arabic emotion analysis by aspect rather than Visualize all the insights in a customer insights dashboard
Arabic natural language processing (Arabic NLP) powers the sentiment model, such that it differentiates between Arabic dialects while picking up on colloquialisms, language nuances, social media short forms, and even emojis.
Repustate enables you to quickly and accurately capture customer and employee sentiments to increase efficiency and improve customer experience, provides native language analysis for 23 languages, and makes social media listening effortless by seamlessly integrating with the world's most popular social networks, review sites, and news sources.
Watson NLU
IBM Watson is one of the most well-known conversational AI platforms.
IBM Watson Natural Language Understanding gives you access to detailed developer resources that help you get started fast, including documentation and SDKs on GitHub.
The Arabic Natural Language Understanding enables users to extract meaning and metadata from unstructured text data. Text analytics can be used to extract categories, classifications, entities, keywords, sentiment, emotion, relationships, and syntax from your data.
Some high-level features of the platform
- Train Watson to understand the language of your business and extract customized insights with Watson Knowledge Studio.
- Surface real-time actionable insights to provide your employees with the tools they need to pull meta-data and patterns from massive troves of data.
- Deploy Watson Natural Language Understanding behind your firewall or on any cloud.
There are some Arabic language limitations, some features are not supported in Arabic such as classifications, concepts, emotions, and semantic roles for these features.
Azure Cognitive Service
Azure Cognitive Service for Language is a new cloud-based service that provides NLP features for understanding and analyzing text.
This language service unifies Text Analytics, QnA Maker, and LUIS and provides several new features.
Most importantly it supports 96 languages including Arabic.
You can create an FAQ bot trained on unstructured data or use this to create advanced conversational experiences with the Microsoft Bot Framework.
Other Options
This is not an exhaustive list. There are many other Arabic NLP options out there (e.g Farasa, MADAMIRA, and Stanford (CoreNLP)
Botpress
Botpress is a favourite of ours as it's an all-in-one conversational AI platform.
Most importantly for this post is that the Botpress natural language understanding engine also provides Arabic natural language understanding out of the box.
Botpress is a platform that makes it easier for developers to create chatbots.
The platform assembles all of the boilerplate code and infrastructure you'll need to get a chatbot up and running, as well as providing a complete dev-friendly platform with all of the tools you'll need.
The platform contains the following features:
- To build multi-turn conversations and workflows, there's a visual Conversation Studio.
- To simulate chats and debug your chatbot, you'll need an emulator and a debugger.
- Natural Language Processing activities are built-in, including intent categorization, spell checking, entity extraction, and more.
To expand the functionality, there is an SDK and a Code Editor.
Botpress is multi-channel so your Arabic chatbot can be deployed to Slack, Telegram, Microsoft Teams, Facebook Messenger, and an embeddable online chat are among the major messaging services supported.
The platform also provides Analytics, human handoff, and other post-deployment technologies.
Botpress facilitates the creation of FAQ-style chatbots. Typically, this chatbot will rely primarily on pre-populated responses.
The platform also enables you to create more complex multi-turn conversational experiences capable of comprehending Arabic and communicating in a human-like manner. They may extract information like dates, amounts, and locations from talks.
Botpress, like any other adaptable chatbot builder platform, offers limitless bot development possibilities. Botpress may be used for almost anything, from virtual enterprise assistants to consumer-facing bots that live on popular messaging networks.
Botpress Interface Features
Although it's beyond the scope of this document to review the Botpress platform in too much detail it's useful to briefly cover the basics.
The first thing that should be mentioned is that the interface of the platform is very smooth and easy to learn in a short time, building a chatbot using Botpress is quite simple, Let's review the interfaces of Botpress.
Studio Interface
When you choose a bot, you'll be taken to the Conversation Studio. For a new chatbot, Conversation Studio creates a new flow. Update the conversational flow and train an NLU model after testing, and then test and debug the chatbot Flows
Using a user-friendly design, the Flows page assists you in creating a conversational flow.
Natural Language Understanding
Botpress is an intent-based platform. You can create intents and train the model with utterances and specify how the bot should respond. The platform also offers many of the standard NLP features:
- Entity extraction. Every phrase contains entities that help your bot understand a user’s intent and respond appropriately.
- System and custom entities. System entities are known entities that you can incorporate into your bot to accelerate development. You can also provide custom entities in the form of patterns or lists.
- Slots. These are the parameters that must be fulfilled to complete an action associated with intent. You define your slots and the NLU tags certain words from a user input that can be identified as intent slots.
- Slot filling. The engine gathers info required to satisfy a particular intent.
Q&A
The user can post frequently asked questions and their answers using the Q&A page.
Libraries
You can use hooks and actions on the Libraries page to import your custom code.
Analytics
The Analytics page shows dashboards that contain analytics data obtained during user chats.
Bot Improvement
The Bot Improvement tab helps you to monitor and develop your chatbot by managing negative comments from users.
Other Features
- Broadcast: You can use the Broadcast page to deliver information to a big group of individuals.
- Code Editor: Without leaving the Botpress Conversation Studio, you may create and update actions, hooks, libraries, configurations, and module configurations on the Code Editor page.
- HITL Next: The HITL page allows you to integrate humans into the loop of the conversation when human intervention is needed.
- Misunderstood: The Misunderstood page includes the user's input that triggered the error-handling cycle, as well as when they give negative feedback regarding the Q&A.
- Testing: You can build conversation scenarios on the Testings tab to confirm that the bot maintains its good behaviour regardless of the scenario. Unit tests are what they're called.
Arabic Chatbot POC
The intention is to build an Arabic Chatbot by using the Botpress platform which supports the Arabic language.
Botpress was chosen for this project because the easy-to-use interface and out-of-the-box functionality allowed us to create a working chatbot fairly quickly.
For this project, it's going to be an Information Provider only for a Hotel chatbot concierge. A simple FAQ Bot which is the customer will ask and the bot will respond. We used the Q&A feature in Botpress to train the bot in Arabic to understand and respond to questions.
The challenge that was faced in the early stages was that there is not enough information about the Arabic language that may help to build the best Chatbot. There is scope for more information.
Tips. Insight. Offers. Are You In?
Conclusion
There are a number of excellent natural language tools and conversational AI platforms available to create chatbots that can converse in Arabic, with the accuracy and technology of Arabic natural language understanding improving day by day.
However, there are still challenges in creating and maintaining Arabic chatbots. This is compounded by a skills shortage of Arabic speakers in the AI world who have experience in creating chatbots in multiple languages and dialects and designing conversations in these languages whilst taking each nuance of a specific language into account.
Natural Language Processing (NLP) is a challenging field and it feels like some of the major players in this space need to step up their game. Google Dialogflow and Amazon Lex are conspicuous in their absence of Arabic support.
Of course, even if Arabic NLU's strength has increased significantly, it is always possible to improve it. The NLU engines are improving all the time, and further breakthroughs are undoubtedly on the way. There will always be work to do until NLU reaches anywhere near human levels.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
9 Questions To Help Define Your Chatbot Project Scope [2023 Update]
Defining Your Chatbot Project Scope
So what is a scope of work?
In terms of conversational AI development, a scope of work typically outlines the specific tasks and objectives that will be accomplished during the project. This can include details such as the functionality of the chatbot, the technologies and platforms to be used, the process for testing and deployment, and any ongoing maintenance and support that will be provided.
So it is essentially a detailed description of the project's deliverables and serves as a blueprint for the development team to follow. The scope of work is an agreement between the client and the development agency that outlines what will be delivered and the expectations of both parties in terms of cost, timelines and quality.
The success of a chatbot or voice assistant project relies heavily on the initial planning and discovery stages. It is crucial to establish strong foundations during these stages in order to ensure a smooth progress throughout the project
Chatbot and voice assistant projects can vary greatly in terms of scope and complexity. At The Bot Forge, we often receive inquiries from customers with a wide range of conversational AI project ideas. Some have already done extensive research and have a clear set of high-level requirements in the form of use cases or user stories. Others, however, may only have a general understanding of their problem and want to explore the potential of a chatbot or voice assistant as a solution
By clearly defining the scope of the project, it will be easier to identify the resources, timelines, and budget required to successfully deliver the project
Asking the right 9 questions
Scoping a chatbot or voice assistant project is largely about asking the right questions. We often find the questions we need to ask to pin down the high-level requirements and get an idea of scope are often the same. So we try to get answers to the following questions as soon as possible:
Tell us about your business/department/area...
e.g. We are the market leaders in creating x. My team handles customer service for our organization. We have x staff members and handle approximately y queries a day. Our key metrics are z and z.
What are the business problems that you'd like to address?
e.g. “we have a problem in our company, with getting too many questions about x via y, and they are often the same. What can a chatbot do to help us? Where can it be used?"
What are your objectives?
e.g. we want to use a chatbot to answer these repetitive questions so our team can concentrate on answering the more difficult queries
What is the expected process flow and/or user journey of the chatbot?
e.g. Our users request information via phone call/email/IM. 80% of these questions are simple and repetitive. We want to address these key queries via a chatbot. Further, we wish to provide our users with a simple way to speak to a live agent if more information is needed.
What are the use cases that the chatbot or voice assistant needs to satisfy and how complex are they?
e.g. we have a handful of questions which we get asked a lot. These are standard customer support requests such as where is my order? These can be answered in a few minutes. We also receive 50 or 60 other inquiries about our products which can be very detailed.
What channels does the chatbot need to be deployed to?
e.g. the chatbot will need to be deployed onto our department website as a web widget and a Microsoft Teams chatbot.
Do you need to support hand over to live agents to answer advanced customer queries or to step in when needed?
e.g. there are live agents which we would like to hand over the chat conversation to
What are the additional requirements and/or conditions?
e.g. we need to log a ticket detailing the chat into our system. We need a Spanish language version
How are you going to measure success?
e.g faster response to the customer, reduced hours spent answering questions, NPS
Once we have the answers to these questions, businesses invariably want a rough estimate: how long will the solution take to build? How much will that cost? We find we are now in a better place to give some high-level estimates. We are also in a position where we understand what our customer's problems are and how we can address them with conversational AI.
TIP: Download our free chatbot checklist to help you iron out your scope
Digging into more detail for chatbot project scope
Building on our standard questions it's then time to look in more detail at your scope:
- Purpose: Clearly define the purpose of the chatbot, such as automating customer service, providing information, or completing transactions.
- Functionality: Identify the specific tasks that the chatbot will be able to perform, such as answering frequently asked questions, booking appointments, or processing orders.
- User flows: Define the user flow and interactions that the chatbot will have with the user, including the type of input and output (text, voice, etc.).
- Integrations: Identify any systems or APIs that the chatbot will need to integrate with, such as a CRM or inventory management system.
- Languages and tone: Define the language and tone that the chatbot will use to communicate with users.
- Data and security: Define the data that the chatbot will access and the security measures that will be put in place to protect that data.
- Channel Deployment: Identify the platforms and channels where the chatbot will be deployed, such as a website, mobile app, or messaging platform.
- Ongoing maintenance and support: Identify the ongoing maintenance and support that will be required to keep the chatbot running smoothly, such as bug fixes and updates.
- Performance metrics: Define the performance metrics that will be used to measure the chatbot's success, such as response times and customer satisfaction.
- Timeframe: Define the project's start and end dates and the milestones that will be reached along the way.
Tips. Insight. Offers. Are You In?
You Have Your Scope... What Next?
Once you have your chatbot scope it’s time to work closely with your chatbot solutions provider (*tip*, that’s The Bot Forge!) in a more detailed project planning phase which can be broken down into the following steps:
- Sure, here are the steps for the project planning phase of a conversational AI project such as a chatbot or voice assistant:
- Define project goals, and objectives: Clearly identify the project's purpose, what it will deliver and what it will not deliver. Define specific, measurable, achievable, relevant and time-bound (SMART) goals to align project objectives with the overall business objectives.
- Identify stakeholders and establish communication plan: Determine who will be impacted by the project, who will be involved in the project, and who will be responsible for approving and implementing the project. Create a communication plan to keep stakeholders informed of project progress, decisions, and issues.
- Assemble project team and assign roles and responsibilities: Identify the resources required to complete the project, including internal team members and external vendors or contractors. Assign specific roles and responsibilities to team members to ensure clear accountability and effective collaboration.
- Create a project plan and schedule: Develop a detailed project plan that outlines the project's activities, dependencies, milestones, and deliverables. Create a schedule that shows the start and end dates for each activity, and allocate resources accordingly.
- Identify risks and develop a risk management plan: Identify potential risks that could impact the project's success, such as budget constraints, schedule delays, or resource shortages. Develop a risk management plan to mitigate or avoid these risks.
- Obtain approval and funding: Once the project plan is complete, present it to stakeholders for review and approval. Obtain the necessary funding to support the project.
- Implement the project plan: With the project plan approved, project team can start to implement the plan and execute the project activities as per the schedule.
- Monitor and control the project: Continuously monitor the project's progress against the plan, identify and resolve issues, and make adjustments as needed.
- Close the project: Once the project is completed, document the results, and perform a post-project review to identify what worked well, what could be improved, and what lessons were learned.
Requirements analysts, Product managers, and stakeholders, conversation designers, AI trainers and developers all need to be involved in the planning stages. The key to a successful project is to ensure the whole team has a shared vision of what the problem is, what the solution is, and who the solution is for. It's this mutual understanding which will ensure the successful creation of a high-level plan and the bedrock for a successful conversational AI project.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
What’s The Difference Between ChatGPT & GPT3?
ChatGPT Confusion
At the time of writing, since the launch of ChatGPT at the end of November 2022, numerous solutions have hit the market claiming to be ChatGPT-branded. However, I'm here to clarify that these solutions are not ChatGPT, but rather GPT3 solutions.
There seems to be a lot of confusion between ChatGPT and GPT3.
This is compounded by some solutions in the marketplace claiming to use ChatGPT which is either totally wrong, trying to ride the hype train, or genuinely confused with the underlying tech.
So, What's The Difference Between ChatGPT & GPT3?
If you got this far, you're probably wondering, what the differences are between ChatGPT and GPT3.
GPT3 stands for Generative Pre-trained Transformer, which is a Large Language Model (LLM) built by OpenAI and released in June 2020. The GPT3 model was later iterated into GPT3.5, also known as InstructGPT, to improve its ability to follow instructions and complete tasks.
If you are using the Davinci model from OpenAI's API, you are using InstructGPT or GPT3.5.
Many companies have already used GPT3 and GPT3.5 to enhance their existing products and create new ones, such as AI-assisted writing tools. However, these are not ChatGPT.
ChatGPT has undergone further training, including RLHF (Reinforcement Learning from Human Feedback).
The training process involved humans reviewing and rewriting responses generated by the model to make them more factually accurate and conversational.
The responses were then fed back into the model to train it on how to produce more human-like answers.
The model then underwent a reward model training process where multiple responses were generated and ranked by humans based on their quality and fit.
The data was then fed back into the model to train it on what constitutes a good response.
OpenAI also used Proximal Policy Optimization (PPO), a Reinforcement Learning approach, to create policies for the ChatGPT language model.
This process not only improved the accuracy and conversational breadth of responses but also trained the model to produce safer content by blocking racist, sexist, or inappropriate responses.
However, despite being built on the GPT3.5 model, ChatGPT produces better responses than GPT3.5, as it has been specifically trained to generate high-quality, human-like responses.
This is why people are often confused between the two models.
What Can You Do With GPT3?
GPT3 is incredibly powerful and ultimately the foundation of ChatGPT. Its applications are wide-ranging because it has an API to build plugins or tools that use Large Language Models.
It can generate text, answer questions, translate languages, summarize long articles, and even complete tasks such as coding and creating charts.
Here are a few things you can do with GPT-3:
- Text generation: GPT-3 can be used to generate human-like text in various styles, from creative writing to news articles, and even poetry
- Question answering: GPT-3 can answer a wide range of questions with high accuracy, making it a useful tool for knowledge management and customer service
- Chatbots: GPT-3 can be used to develop advanced chatbots that can handle complex conversational tasks, such as booking a flight or ordering food
- Language translation: GPT-3 can translate text from one language to another, providing near-human-level accuracy
- Content summarization: GPT-3 can summarize long articles or documents into concise summaries, making it easier to quickly understand the most important information
- Code generation: GPT-3 can write code, from simple scripts to complete applications, making it a valuable tool for software development
- Creative applications: GPT-3 can be used for creative projects, such as generating music, visual arts, or even video game design.
What Can You Do With ChatGPT?
As I've covered previously ChatGPT is built to have conversations and do it well. Its conversational capabilities (remember only its own chat interface) are better than GPT3 because it's been tuned to do it. It understands natural language input and generates human-like responses to questions and supports follow-up questions.
Some of the things that can be done with ChatGPT include:
- Question answering: ChatGPT can answer a wide range of questions, providing users with relevant and accurate information on a variety of topics
- Conversation simulation: ChatGPT can mimic human conversations, making it ideal for use in customer service, virtual assistance, and other scenarios where a human-like interaction is required.
- Text generation: ChatGPT can be used to generate text, such as product descriptions, headlines, and other types of content
- Summarization: ChatGPT can be used to summarize long articles, news reports, or other written content into shorter, more concise text
- Translation: ChatGPT can be used to translate text from one language to another, making it a useful tool for global communication.
Summary
So.... ChatGPT is built on the GPT3.5 model but is a conversational interface accessible only through a browser and does not have a publicly available API..at the time of writing. One thing we know about conversational AI is that the landscape is changing quickly and an API release is coming soon.
GPT3.5 is available via an API and a browser and it's this technology that can be used to create lots of different applications including chatbots.
It's also worth mentioning that GPT3.5 can also be leveraged to provide conversational experiences "Like" ChatGPT. Remember they are not ChatGPT, but it's possible to leverage GPT3.5 and add enhanced conversational capabilities which are similar. So memory and context coupled with a smaller domain area knowledge, not the entire internet!
This type of implementation is possible and is something we are working on at The Bot Forge so feel free to get in touch if you'd like to learn more.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
What is ChatGPT?
AI-Mazing
ChatGPT is the latest technology release from the team at OpenAI and it's taken the internet by storm. Reactions ranged from amazement to scepticism, with everything in between.
It's been hugely popular with more than a million people using it in the first 5 days.
ChatGPT is OpenAI’s latest large language model, released on Nov 30th 2022, in beta as a chatbot app which you can interact with on the OpenAI website.
Like GPT3, it can explain scientific and technical concepts in a desired style or programming language, brainstorm basically anything you can think of asking... and yes, of course, hold pretty complex conversations!
Like any new technology in this field, particularly one which has caused such a stir, it's important to not get too over excited and consider the real value, what can it do and what are its limitations.
Let's take a look...
So, What is ChatGPT?
You're probably wondering how it works.
The system is built on a type of artificial intelligence (AI) called a large language model. The ChatGPT model is based on reinforcement learning from human feedback and the GPT series of models from OpenAI which are themselves trained on extraordinarily large amounts of data.
To create ChatGPT, the latest GPT-3.5 Instruct model was fine-tuned with conversation examples, instead of the whole internet, to concentrate on improving the model's specialist conversation abilities. Reinforcement learning was then used so that the model could practice its conversation skills and improve.
The system works like any online chatbot, and you can simply type out and submit any question or prompt you'd like the AI to address.
The ChatGPT service is currently accessible via a web chat interface. Its simple interface hints at what it can do but also some of its limitations.
What Can ChatGPT Do?
ChatGPT can do a hell of a lot. It's beyond the scope of this post, or any post for that matter, to cover everything it can do, but let's focus on some of the things it does well.
It's also worth mentioning that all of these tasks are achieved automatically. No need to write complex prompts, configure temperature, or fine-tune models.
It can have complex conversations with you
The new chat capability of ChatGPT allows for a pretty natural conversation experience.
It is the first chat-focused large language model and it's really good at it.
Previous attempts at creating GPT3-based chat experiences has been hacky and pretty unreliable.
It has long-term memory of up to 8192 tokens and can take input and generate output about twice as long as GPT3.
All of this allows ChatGPT to maintain context and generate follow-up responses with astounding ease.
The ability which ChatGPT has to hold a multi-turn conversation is, in my opinion, the real game changer here.
To create a similar experience using the latest Intent-based conversational AI technology, i.e Rasa, involves a serious amount of work.
It can have safer conversations
In order to make ChatGPT "safe," OpenAI spent a lot of time attempting to disable responses that dealt with violence, terrorism, drugs, hate speech, dating, sentience, and eradicating humanity.
OpenAI also claimed to disable web browsing and knowledge of current dates.
It's still relatively easy to compromise these security efforts though. OpenAI have placed clear warnings that this is a test and encourages reporting of results.
So you can assume this feedback will be used to make service safer.
It's incredibly creative
As we've previously discussed, the best use cases for ChatGPT involve creative writing, brainstorming, drafting, and creative information presentation.
Any tasks where originality is valued more highly than accuracy.
Writing blog posts and content
From blog posts, to summarisation, to copywriting, to writing lecture notes about complex subjects... ChatGPT is pretty good at taking on any of these tasks.
Normally after a little bit of polishing and editing the end result is really impressive.
Writing lyrics and short stories
It's also amazing at generating rhyming poetry and producing lyrics and understanding the components of a song i.e chorus, verse, and bridge.
This is a huge leap compared to GPT3 where, from our experience, you have to finetune a DaVinci model based on a specific lyrics dataset.
It Can Help Create Better Conversational Experiences
ChatGPT and LLMs can do a lot to help create better chatbots and voice assistants.
This is really interesting for the world of conversational AI.
Whilst some people in the industry are looking nervously at their intent-based models, others are imagining a better world where LLMs such as ChatGPT can really help them to create better chatbots and voice assistants.
Here are some examples of what ChatGPT can do:
- Creating intent utterances
As a tool for writing utterances for intents, it's fantastically capable of providing training data for intent matching. In the absence of any available customer conversational data, ChatGPT utterances can serve as a good means to bootstrap a chatbot ready for further iterations.
- Entity generation
Create entities by asking for permutations e.g "list 10 different ways of saying desktop" or "list 10 different types of cycling"
- Prompt variations
It's good practice to create different prompts to make for a better experience, you all knew that right? Well ChatGPT is brilliant for creating different ways of saying the same thing: "give me 5 ways of saying would you like a tip of the day"
- Happy path creation
ChatGPT can help provide some conversation path examples if you provide it with an outline of the chatbot/VUI you are creating. We've been using this to create happy paths for Alexa skills but this approach could be used for any conversational experience.
- Persona creation
ChatGPT can also be used to simulate conversations and provide more detail about specific personas
- Test case generation
This is a really useful one. If you want to create test utterances to train a model then for each specific intent you can simply ask ChatGPT to create some for you. Just explain the intent itself and away you go
What's even more useful is the ability to prompt ChatGPT to create long tail test utterances with a simple follow-up question to the previous one.
It Can Write Code Examples
There are already some really interesting software products based on GPT3 Codex with one of the most standout apps being GitHub Copilot.
ChatGPT continues this with the ability to create pretty good code output. Here is an example of a request for some simple Javascript code. Note the fantastic formatting capabilities of ChatGPT:
ChatGPT even has the ability to create test code "can you create some unit tests in jest to test the object" and even to debug code to a limited degree.
What are ChatGPT's Limitations?
ChatGPT's drawbacks are highlighted nicely on the ChatGPT UI:
- May occasionally generate incorrect information
- May occasionally produce harmful instructions
- Limited knowledge of world and events after 2021
This is a pretty good summary.
ChatGPT Sounds like many of the millions of people posting on Twitter; they sound confident but can still be wrong.
The security systems which are obviously a large part of OpenAI's roadmap for the technology can still be fairly easily broken.
Yes, it's a large language model which has only been trained on historical data so there is a limitation there if you want to provide information.
However, ChatGPT is aware of its own limitations. If you ask it about the current situation in Ukraine:
"I'm sorry, but I do not have current information about the situation in Ukraine. My training data only goes up until 2021."
What's Next For ChatGPT?
We feel there is still a lot to come. Definitely including improved safety and we feel there are more features to come. There have already been features added including conversation history and improved performance.
The brilliant formatting capabilities mean ChatGPT could be moving towards the ultimate encyclopedia, although the risks of misinformation are still there.
The costs of a technology like this are eye-watering, so we're pretty sure that at some point ChatGPT is going to cost. There are already signs of this from the daily platform limits being introduced in the latest release.
There is no doubt that there is some big money here. At the time of writing, ChatGPT has been valued at $9 Billion, and OpenAI's valuation may be approaching $30 Billion.
It remains to be seen how much OpenAI will charge for ChatGPT when it's finally released into production.
Conversational Search
There has been a lot of discussion about conversational search products and whether they will soon rival the big search engines. Many people immediately saw ChatGPT as superior to Google search. However the reality is that they are very different. The biggest points to make are that the information from ChatGPT is often not correct, responses fabricated and not up to date.
At the moment there are a lot of possibilities for conversational search and large language models and it's likely that the big search engine providers will be focusing on this area over the course of 2023.
A note on watermarking
The quality of the output from ChatGPT has already led to students and others passing off ChatGPT as human-generated. This is being treated as seriously as plagiarism.
Because of the rise in "AIgiarism" (AI-assisted plagiarism) there are increased calls for ways to identify where AI is used to generate content.
Metadata watermarking looks like a viable option to combat this.
There are already OpenAI detector tools on Huggingface. OpenAI has a working prototype of the watermarking scheme that “seems to work pretty well", according to an OpenAI researcher.
It's suggested that a few hundred tokens – or a paragraph of text – is the point needed to get a reasonable signal that the text came from GPT3.
There is going to be more demand for this sort of technology to tackle Algiarism, mass propaganda generation, or writer impersonation.
Conclusion
It's pretty mind-blowing and I've really only scratched the surface of what's possible in this post.
ChatGPT and LLMs look set to change the conversational AI landscape forever.
If OpenAI couples the ability to fine-tune models based on ChatGPT with your own knowledge bases then this would enable the creation of a conversational FAQ with ChatGPT's engaging conversational abilities which would really open up some fantastic possibilities.
There will undoubtedly be an increase in the use of intent based and LLM conversational AI experiences.
And finally, it's really important to note that, at the time of writing, the ChatGPT service is in beta, and not production ready. It's only available for human use, via the OpenAI UI.
There is no API to use to talk to ChatGPT, unlike the other OpenAI models which are easy to use and integrate into existing workflows and tools.
For now, sit back and figure out your next question for ChatGPT.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.
Build a Custom AI-Powered GPT3 Chatbot For Your Business
The Rise Of ChatGPT
ChatGPT has rapidly gained popularity worldwide, with millions of users relying on its vast knowledge database and ability to hold a multi-turn conversation of considerable complexity.
However, despite its usefulness for general information, ChatGPT is limited to pre-2021 publicly available internet data, and it has no access to your private data or recent sources of information. Imagine how beneficial it would be to your business if something like ChatGPT had access to this information.
There has been considerable demand for chatbots similar to ChatGPT but without these limitations. Let's take a look at what is possible.
What is a Large Language Model (LLM)?
A large language model is a type of artificial intelligence model that is designed to understand and generate human language.
It's built using machine learning techniques and is trained on massive datasets of natural language text to learn the patterns and rules of language. These models use deep learning algorithms, such as neural networks, to understand the relationships between words, phrases, and sentences, allowing them to generate text that sounds natural and human-like.
One of the most well-known examples of a large language model is GPT-3 (Generative Pre-trained Transformer 3), which was developed by OpenAI and has become one of the most powerful language models available.
GPT-3 is particularly powerful because it has been trained on an enormous amount of data - over 570GB of text - making it one of the most sophisticated language models currently available.
Other Large Language Models Are Available
Although we've only referred to the GPT-3 large language models provided by OpenAI, it's worth noting that there are a number of other LLMs. Each LLM has its own strengths and weaknesses when asked to process and understand natural language in various ways; here are some examples:
- T5: T5 (Text-to-Text Transfer Transformer) is a large-scale language model developed by Google. It was trained on a diverse range of text-based tasks and can perform various natural language processing (NLP) tasks such as text summarization, question-answering, and translation.
- BERT: BERT (Bidirectional Encoder Representations from Transformers) is another large-scale language model developed by Google. It can be used for various NLP tasks, including question-answering, sentiment analysis, and text classification.
- RoBERTa: RoBERTa (Robustly Optimized BERT Pretraining Approach) is a large-scale language model developed by Facebook AI. It is an improvement over BERT and performs better on several NLP benchmarks.
- XLNet: XLNet is another language model developed by Google that uses an autoregressive approach for language modelling. It achieves state-of-the-art performance on several NLP benchmarks.
- GShard: GShard is a distributed large-scale language model developed by Google that achieves state-of-the-art performance on several NLP benchmarks. It is trained using a novel hierarchical approach that enables it to scale to trillions of parameters.
- Bloom: Bloom, said to be the world's largest open multilingual language model, is one of the latest LLMs and is available via the Hugging Face platform.
Access to these models is also available via other providers e.g Cohere, GoogleAI, and Hugging Face.
The takeaway here is that LLM technologies are becoming increasingly accessible to create tailored conversational experiences.
Are we using ChatGPT or GPT-3?
It's easy to get confused between ChatGPT and GPT-3, something we've looked at before in detail.
GPT stands for Generative Pre-trained Transformer, which is a Large Language Model (LLM) built by OpenAI and released in June 2020. The GPT3 model was later iterated into GPT3.5, also known as InstructGPT, to improve its ability to follow instructions and complete tasks.
What makes GPT-3 so groundbreaking is its ability to generate natural language text that is virtually indistinguishable from text written by humans. The model is trained on an incredibly large dataset of internet text, including books, articles, and websites, which allows it to understand the nuances of human language and generate responses in a natural, conversational style.
On the other hand, ChatGPT is built on top of GPT3 but has been enhanced with further training processes.
What we are going to be looking at in this post is using GPT-3 to create an experience similar to or like ChatGPT... but not using ChatGPT!
Use Your Data To Power a GPT3 Chatbot
Part of the challenge of creating a large language model chatbot based on your organisation's data is accessing the data and loading it into the correct form for it to be used in the data ingestion process.
Increasingly, businesses store their knowledge in various locations, depending on the type of knowledge and the company's specific needs and across a range of formats. However, more often than not this unstructured text data will be in a form that you can work with. Some good examples include:
- HTML
- PowerPoint
- Podcast content
- YouTube video transcripts
- Internal databases
- Customer support queries
- Other APIs
- Documentation sources e.g GitBooks
Once a data source has been identified and extracted, the next stage is to clean and preprocess the data to ensure that it is in a format that can be used. This process may involve removing duplicates, cleaning and labelling text, and standardizing formatting to ensure consistency across different data sources.
This data can then be processed and used in your interactions with your large language model.
Technologies To Interface With LLM
Whilst it's entirely possible to code up a solution to interact with LLMs from the ground up that is also time-consuming and complex. There is a growing list of offerings that can help achieve your conversational AI use-case goal.
The technologies for orchestrating chatbots based on LLMs like GPT3 are evolving rapidly. These technology stacks provide the tooling we need to create a conversational engine that can interact with LLMs easily.
Orchestration
Functionality across the different platforms falls into the same categories of existing conversational AI platforms with offerings falling into the classes of Pro-Code, Low-Code and No-Code solutions.
There are a number of these tools/platforms currently available e.g Dust, Langchain - each could warrant a dedicated post.
It's a bit of an oversimplification of what these technologies actually do but as a summary, they provide the features needed to carry out the steps needed to create conversational use-cases such as chatbots, text generation, and Q&A by interacting with a LLM.
Features
For our use case, we are looking to create a conversational agent similar to ChatGPT so the following features all come into play:
- Tools to make sense of large volumes of unstructured text data
- Tools to work with Vector stores
- Prompt generation assistance (A prompt is an input to a language model, a string of text used to generate a response from the language model).
- Accessible wrapper to talk to your LLM of choice
- Tools to enable the management of conversation state and context
Vector Stores
A vector store is a specific type of database optimized for storing documents, and embeddings, and then allowing for fetching the most relevant documents for a particular query.
These are important for our GPT3 knowledge-base powered chatbot as they store our document embeddings as indices for the search. Notable libraries are the FAIIS open-source library and the Weaviate open-source vector search engine
Creating a GPT3 Chatbot
For this example, we'll look at using Langchain to create our GPT3 chatbot. Of the platforms mentioned earlier, Langchain is our favourite. It's pro-code but is well-supported with examples and documentation.
We've included high-level technical detail in this guide; here are the main steps.
- Source your data: Take your unstructured text and clean it and prepare it for use.
- Chunk/Embed text/Load embeddings: Load your text into smaller pieces. Convert each chunk of text into a numerical format so that you can find the most relevant chunks for a given question. Put the numerical embeddings and documents into a vector store, which helps to quickly find the most similar chunks of text to a given question.
- Prompt engineering: Create the correct prompts to pass onto the LLM based on context, question history and required behaviour.
- Deploy your chatbot: Integrate the chatbot service into your channel, this could be really simple... or a slick chat UI similar to ChatGPT.
- Talk to your chatbot: Once deployed, you can start asking questions. When a user submits a question, the chatbot will identify the most relevant chunks of text in the vectorstore and generate a response based on that information and the current context and conversation state.
- Fine-tune the model: As the chatbot is used, you may find that it is not always generating the best responses. In that case, you can fine-tune the model by adding more data or adjusting the weighting of different chunks of text in the vectorstore.
- Evaluate the performance: To ensure that the chatbot is working as intended, you should regularly evaluate its performance. This may involve analyzing user feedback, monitoring the accuracy of its responses, or testing it against a range of different queries and of course keeping your vectorstore up to date with any new information.
What Is The Result?
Results
As a very quick POC, we ingested all the text from thebotforge.io and ran it through our process. The results are actually pretty good.
A GPT3 chatbot project created using the latest LLM orchestration stack provides good results on an unstructured dataset e.g website contents. It handles context, so follow-up questions about the ingested data work pretty well.
We can ask specific questions about the subject matter in a number of different ways, and ask follow-up questions.
To be honest, it's not as good a ChatGPT, but we wouldn't expect it to be. Its capabilities aren't as wide-ranging, but that is perhaps the point - they don't need to be. The main capability is that it provides the ability to talk about your knowledge base with much more flexibility than an intent-based conversational AI experience, which would take a lot longer to create and would be unlikely to be anywhere near as powerful.
Hallucinations
LangChain helps to overcome hallucinations which is an issue with LLMs.
In the context of large language models (LLMs), "hallucinations" refer to when the model generates text that is not coherent, relevant, or accurate. Hallucinations can occur when the LLM generates text that is not based on the input or task at hand but is instead based on its own learned patterns or biases.
This can happen because the LLM has learned certain patterns in the data that do not apply to the specific context of the task. In other cases, an LLM may generate text that is completely unrelated to the input or task, which can be described as "hallucinating" text.
To mitigate the risk of hallucinations, LLMs need to be trained on high-quality data, and the generated text needs to be evaluated to ensure that it is relevant and accurate to the task at hand. This can be handled within the conversation itself e.g highlighting knowledge-base content with chat responses which is where tools like Langchain come in.
Limitations
OpenAI Services
There are limitations related to the OpenAI service.
One is the cost of interacting with OpenAI's models. In the case of using OpenAI for our Langchain example, this could get expensive pretty quickly as we are using Davinci 003 which is the most capable of their current models, but also the most expensive. We also found we are running close to the maximum prompt size for our interactions.
The second issue is that we found the API calls can be laggy at times, which means poor performance for the chat interface, and more worryingly we received rate limit errors from the service because of high traffic.
Transactional Chat
No intents and integrations here. So where the ability to handle free conversation is good if a user of your GPT3 chatbot is at a stage of a conversation where they need to carry out a specific task, then this is where your chat service would need to hand over to a more intent-based approach.
We've found that a blended approach of LLM & Intent-based service works well here. Catch a support intent from a user then hand them over to your conversational AI intent service or live chat agent to manage the transaction... you can even hand them back once it's complete.
Conclusion
There is no doubt that ChatGPT has gained a huge amount of traction over a short space of time, but it's worth remembering that it's based on GPT3 technology which has been around for a while.
Despite the wonder of ChatGPT's ability to follow a line of conversation any number of times about any number of subjects, it still has obvious limitations. The most notable being there is no API (at the time of writing), it's trained on data up to 2021 and it has no real knowledge of your organisation's recent or private data.
Let's not forget that the essence of any LLM's functionality is to produce a reasonable continuation of whatever text it's got so far. It's ChatGPT's conversational "qualities" which you could argue have driven its popularity.
To handle questions about a subject or domain specific to your organisation then a GPT3 or other LLM-powered chatbot makes a lot of sense, particularly when you can give it similar "qualities" to ChatGPT.
Overall the future looks bright for LLM-powered conversational experiences. Technology in this space is progressing rapidly with a ChatGPT API in the pipeline and with rivals to ChatGPT already planned e.g Hugging Faces' next version of the BLOOM LLM.
It's also going to be the continued advancements in smaller scale fine-tunable streamlined LLMs and automated NLP model compression and optimisation tools which will begin to power a lot of our chatbot conversations.
If you want to talk to us about leveraging AI and your organisation's data, get in touch.
About The Bot Forge
Consistently named as one of the top-ranked AI companies in the UK, The Bot Forge is a UK-based agency that specialises in chatbot & voice assistant design, development and optimisation.
If you'd like a no-obligation chat to discuss your project with one of our team, please book a free consultation.