raktimhugging commited on
Commit
602d30e
Β·
verified Β·
1 Parent(s): 6783325

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +76 -210
README.md CHANGED
@@ -1,226 +1,92 @@
1
- # πŸ€– Raktim Mondol's Portfolio with RAGtim Bot
2
-
3
- A modern, responsive portfolio website featuring an intelligent AI assistant powered by multiple backend options including Hugging Face Transformers.
4
-
5
- ## 🌟 Features
6
-
7
- ### Portfolio Website
8
- - **Responsive Design**: Beautiful, mobile-first design with dark/light theme support
9
- - **Interactive Sections**: Education, Experience, Skills, Research, Publications, Awards
10
- - **Audio Summaries**: Listen to AI-generated summaries of research papers
11
- - **Thesis Mind Map**: Interactive visualization of PhD research (desktop only)
12
- - **Contact Form**: Direct email integration with validation
13
-
14
- ### RAGtim Bot - AI Assistant
15
- - **Multiple Backend Options**: Choose from Hugging Face Space, Backend Server, or Netlify Functions
16
- - **Advanced Search**: Hybrid semantic + keyword search for accurate responses
17
- - **Comprehensive Knowledge**: Covers all aspects of Raktim's expertise
18
- - **Real-time Chat**: Instant responses with conversation history
19
- - **Performance Optimized**: Caching and efficient processing
20
-
21
- ## πŸš€ RAGtim Bot Deployment Options
22
-
23
- ### Option 1: Hugging Face Space (Recommended) πŸ€—
24
- **Pros**: Free GPU acceleration, no API keys needed, automatic scaling
25
- **Best for**: Public deployment, maximum performance
26
-
27
- ```bash
28
- # Set environment variable
29
- VITE_USE_HUGGING_FACE=true
30
- ```
31
-
32
- **Live Demo**: [RAGtim Bot on Hugging Face](https://huggingface.co/spaces/raktimhugging/ragtim-bot)
33
-
34
- ### Option 2: Backend Server πŸ–₯️
35
- **Pros**: Full control, local processing, custom configurations
36
- **Best for**: Development, private deployments
37
-
38
- ```bash
39
- # Set environment variables
40
- VITE_USE_BACKEND=true
41
- VITE_BACKEND_URL=http://localhost:3001
42
- VITE_DEEPSEEK_API_KEY=your_api_key
43
-
44
- # Start backend server
45
- cd server
46
- npm install
47
- npm run dev
48
-
49
- # Start frontend
50
- npm run dev
51
- ```
52
-
53
- ### Option 3: Netlify Functions ⚑
54
- **Pros**: Serverless, automatic scaling, integrated with Netlify
55
- **Best for**: Simple deployments, cost-effective
56
-
57
- ```bash
58
- # Set environment variable
59
- VITE_DEEPSEEK_API_KEY=your_api_key
60
- # (VITE_USE_HUGGING_FACE and VITE_USE_BACKEND should be false or unset)
61
- ```
62
-
63
- ## πŸ› οΈ Quick Start
64
-
65
- ### 1. Clone and Install
66
- ```bash
67
- git clone <repository-url>
68
- cd portfolio
69
- npm install
70
- ```
71
-
72
- ### 2. Choose Your RAG Backend
73
-
74
- #### For Hugging Face Space (Easiest):
75
- ```bash
76
- echo "VITE_USE_HUGGING_FACE=true" > .env
77
- npm run dev
78
- ```
79
-
80
- #### For Backend Server:
81
- ```bash
82
- echo "VITE_USE_BACKEND=true" > .env
83
- echo "VITE_BACKEND_URL=http://localhost:3001" >> .env
84
- echo "VITE_DEEPSEEK_API_KEY=your_api_key" >> .env
85
 
86
- # Start both frontend and backend
87
- npm run dev:full
88
- ```
89
 
90
- #### For Netlify Functions:
91
- ```bash
92
- echo "VITE_DEEPSEEK_API_KEY=your_api_key" > .env
93
- npm run dev
94
- ```
95
 
96
- ### 3. Open Browser
97
- Visit `http://localhost:5173` to see the portfolio with RAGtim Bot!
98
-
99
- ## πŸ“Š RAGtim Bot Comparison
100
-
101
- | Feature | Hugging Face | Backend Server | Netlify Functions |
102
- |---------|-------------|----------------|-------------------|
103
- | **Setup Complexity** | ⭐ Easy | ⭐⭐⭐ Complex | ⭐⭐ Medium |
104
- | **Performance** | ⭐⭐⭐ Excellent | ⭐⭐⭐ Excellent | ⭐⭐ Good |
105
- | **Cost** | πŸ†“ Free | πŸ’° Server costs | πŸ’° Function costs |
106
- | **Scalability** | ⭐⭐⭐ Auto | ⭐⭐ Manual | ⭐⭐⭐ Auto |
107
- | **Customization** | ⭐⭐ Limited | ⭐⭐⭐ Full | ⭐⭐ Medium |
108
- | **API Key Required** | ❌ No | βœ… Yes | βœ… Yes |
109
-
110
- ## 🎯 RAGtim Bot Capabilities
111
-
112
- ### What You Can Ask:
113
- - **Research**: "What is Raktim's research about?"
114
- - **Publications**: "Tell me about BioFusionNet"
115
- - **Skills**: "What programming languages does he know?"
116
- - **Experience**: "Where has he worked?"
117
- - **Education**: "What degrees does he have?"
118
- - **Statistics**: "What statistical methods does he use?"
119
- - **Contact**: "How can I reach Raktim?"
120
-
121
- ### Technical Features:
122
- - **Semantic Search**: Understanding context and meaning
123
- - **Keyword Matching**: Exact term matching for precision
124
- - **Conversation Memory**: Maintains chat history
125
- - **Response Caching**: Fast repeated queries
126
- - **Error Handling**: Graceful fallbacks and error messages
127
-
128
- ## πŸ”§ Development
129
-
130
- ### Frontend Development
131
- ```bash
132
- npm run dev # Start development server
133
- npm run build # Build for production
134
- npm run preview # Preview production build
135
- ```
136
 
137
- ### Backend Development (if using backend server)
138
- ```bash
139
- cd server
140
- npm run dev # Start backend in development
141
- npm run start # Start backend in production
142
- npm run install-models # Pre-download AI models
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
143
  ```
144
 
145
- ### Full Stack Development
146
- ```bash
147
- npm run dev:full # Start both frontend and backend
148
- npm run build:full # Build both frontend and backend
149
  ```
150
 
151
- ## πŸš€ Deployment
152
-
153
- ### Deploy to Netlify (with Hugging Face)
154
- 1. Set `VITE_USE_HUGGING_FACE=true` in Netlify environment variables
155
- 2. Deploy normally - no additional configuration needed!
156
-
157
- ### Deploy to Netlify (with Functions)
158
- 1. Set `VITE_DEEPSEEK_API_KEY` in Netlify environment variables
159
- 2. Netlify Functions will be automatically deployed
160
-
161
- ### Deploy Backend Server
162
- Choose from:
163
- - **Railway**: Easy deployment with automatic scaling
164
- - **Render**: Free tier available, good for development
165
- - **DigitalOcean**: Full control, various pricing options
166
- - **AWS/GCP**: Enterprise-grade with advanced features
167
-
168
- ## πŸ“± Mobile Experience
169
 
170
- - **Responsive Design**: Optimized for all screen sizes
171
- - **Touch-Friendly**: Large buttons and easy navigation
172
- - **Fast Loading**: Optimized assets and lazy loading
173
- - **Offline Fallback**: Graceful degradation when offline
 
174
 
175
- ## 🎨 Customization
176
 
177
- ### Themes
178
- - Light and dark mode support
179
- - Consistent color scheme throughout
180
- - Smooth transitions and animations
 
181
 
182
- ### Content
183
- - Easy content updates via markdown files
184
- - Modular component structure
185
- - Configurable sections and features
186
 
187
- ## πŸ”’ Security & Privacy
188
-
189
- - **No Data Storage**: Conversations are not stored permanently
190
- - **Secure APIs**: All API calls use HTTPS
191
- - **Input Validation**: Prevents malicious inputs
192
- - **Rate Limiting**: Prevents abuse and spam
193
-
194
- ## πŸ“ˆ Performance
195
-
196
- - **Lighthouse Score**: 95+ across all metrics
197
- - **Fast Loading**: Optimized bundle sizes
198
- - **Efficient Caching**: Smart caching strategies
199
- - **CDN Ready**: Optimized for global delivery
200
-
201
- ## 🀝 Contributing
202
-
203
- 1. Fork the repository
204
- 2. Create a feature branch
205
- 3. Make your changes
206
- 4. Test thoroughly
207
- 5. Submit a pull request
208
-
209
- ## πŸ“„ License
210
-
211
- MIT License - see LICENSE file for details
212
-
213
- ## πŸ™ Acknowledgments
214
-
215
- - **Hugging Face**: For providing free GPU-accelerated AI hosting
216
- - **OpenAI/DeepSeek**: For powerful language models
217
- - **Netlify**: For seamless deployment and functions
218
- - **React/Vite**: For the excellent development experience
219
- - **Tailwind CSS**: For beautiful, responsive styling
220
 
221
  ---
222
 
223
- **Live Demo**: [mondol.me](https://mondol.me)
224
- **RAGtim Bot**: [Hugging Face Space](https://huggingface.co/spaces/raktimhugging/ragtim-bot)
225
-
226
- Built with ❀️ by Raktim Mondol
 
1
+ ---
2
+ title: RAGtim Bot - Raktim's AI Assistant
3
+ emoji: πŸ€–
4
+ colorFrom: green
5
+ colorTo: blue
6
+ sdk: gradio
7
+ sdk_version: "4.15.0"
8
+ app_file: app.py
9
+ pinned: false
10
+ license: mit
11
+ ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
12
 
13
+ # πŸ€– RAGtim Bot - Raktim's AI Assistant
 
 
14
 
15
+ An intelligent AI assistant powered by Hugging Face Transformers that answers questions about Raktim Mondol's research, expertise, and professional background.
 
 
 
 
16
 
17
+ ## 🌟 Features
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
18
 
19
+ - **Complete Markdown Knowledge Base**: Loads all portfolio content from markdown files
20
+ - **GPU-Accelerated Search**: Uses `sentence-transformers/all-MiniLM-L6-v2` for semantic similarity
21
+ - **Comprehensive Coverage**: Research, publications, skills, experience, education, statistics
22
+ - **API Endpoints**: Direct access to search and statistics
23
+ - **Real-time Chat**: Interactive conversational interface
24
+
25
+ ## πŸ“š Knowledge Base
26
+
27
+ This Space loads comprehensive information from:
28
+
29
+ - **about.md** - Personal information, contact details, professional summary
30
+ - **research_details.md** - Detailed research projects, methodologies, current work
31
+ - **publications_detailed.md** - Complete publication details, technical contributions
32
+ - **skills_expertise.md** - Comprehensive technical skills, tools, frameworks
33
+ - **experience_detailed.md** - Professional experience, teaching, research roles
34
+ - **statistics.md** - Statistical methods, biostatistics expertise, methodologies
35
+
36
+ ## πŸ” What You Can Ask
37
+
38
+ - Research projects and methodologies
39
+ - Publications with technical details
40
+ - Technical skills and programming expertise
41
+ - Educational background and achievements
42
+ - Professional experience and teaching roles
43
+ - Statistical methods and biostatistics applications
44
+ - Awards, recognition, and professional development
45
+ - Contact information and collaboration opportunities
46
+
47
+ ## πŸš€ API Usage
48
+
49
+ ### Search API
50
+ ```python
51
+ import requests
52
+
53
+ response = requests.post(
54
+ "https://raktimhugging-ragtim-bot.hf.space/api/search",
55
+ json={"query": "What is Raktim's research about?", "top_k": 5}
56
+ )
57
+ results = response.json()
58
  ```
59
 
60
+ ### Stats API
61
+ ```python
62
+ response = requests.get("https://raktimhugging-ragtim-bot.hf.space/api/stats")
63
+ stats = response.json()
64
  ```
65
 
66
+ ## πŸ”§ Technical Details
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
67
 
68
+ - **Model**: sentence-transformers/all-MiniLM-L6-v2
69
+ - **Embedding Dimension**: 384
70
+ - **Search Type**: Semantic similarity with relevance scoring
71
+ - **Knowledge Sections**: 50+ sections across 6 markdown files
72
+ - **GPU Acceleration**: Automatic CUDA detection and usage
73
 
74
+ ## 🌐 Integration
75
 
76
+ This Space can be integrated with:
77
+ - Portfolio websites for intelligent chat assistance
78
+ - Research collaboration platforms
79
+ - Academic networking tools
80
+ - Professional inquiry systems
81
 
82
+ ## πŸ“ž Contact
 
 
 
83
 
84
+ For questions about Raktim Mondol or collaboration opportunities:
85
+ - **Email**: [email protected]
86
+ - **Portfolio**: [mondol.me](https://mondol.me)
87
+ - **Institution**: UNSW Sydney, School of Computer Science & Engineering
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
88
 
89
  ---
90
 
91
+ **Built with**: Gradio, Hugging Face Transformers, PyTorch
92
+ **Powered by**: GPU-accelerated semantic search and comprehensive markdown knowledge base