Sakshi commited on
Commit
7462cd1
·
1 Parent(s): a327219

prompt optimisations

Browse files
policy_analyser/analyse.py CHANGED
@@ -161,6 +161,7 @@ def analyse(file_bytes, end2end = False):
161
  analysis_start = time()
162
  raw_response = call_openai(ANALYSIS_PROMPT + 'Policy : ' + text + f"\n\nConsider today's date as {datetime.today().day}/{datetime.today().month}/{datetime.today().year} for your analysis on waiting periods and dates")
163
  analysis_end = time()
 
164
  print(f'Analysed [{analysis_end - analysis_start}]')
165
  if raw_response is not None:
166
  response.append(
 
161
  analysis_start = time()
162
  raw_response = call_openai(ANALYSIS_PROMPT + 'Policy : ' + text + f"\n\nConsider today's date as {datetime.today().day}/{datetime.today().month}/{datetime.today().year} for your analysis on waiting periods and dates")
163
  analysis_end = time()
164
+ print('Analysis : ', raw_response)
165
  print(f'Analysed [{analysis_end - analysis_start}]')
166
  if raw_response is not None:
167
  response.append(
policy_analyser/prompts/analysis.txt CHANGED
@@ -2,7 +2,7 @@ Given the markdown content of a customer's health insurance policy, analyse the
2
 
3
  Apply the following rules enclosed in triple backticks on the policy to analyse it.
4
  Make sure you are consider values for analysis factors on basis of customer's selected insurance plan when multiple plans are described in the policy terms.
5
- Make sure all factors appear in one of Good, Average or Bad only. No factor should be repeated in more than 1 verdict table
6
  Note : Top cities = [Mumbai, Delhi, Bangalore, Chennai, Hyderabad, Gurgaon, Pune]
7
  ```
8
  IF Adults == 1:
@@ -121,23 +121,33 @@ ELSE:
121
  Verdict = "Bad"
122
  ```
123
 
124
- Format your response in the following way, to present analysis to customer. Use appropriate language and emojis to portray analysis and verdicts to the customer. Be discrete about rules, do not expose rules to customer but use them to explain reasoning and analysis:
125
 
126
  [CUSTOMER_RESPONSE]
127
  # Our Analysis of your policy [Name of policy] by [Name of insurance company]
 
128
  ## Good Factors
 
129
  | Factor | Your policy | Our Analysis |
130
  | --- | --- | --- |
131
- | Sum Insured | Value of sum insred in the policy | Analysis of why sum insured is good based on the given rules |
 
132
 
133
  ## Average Factors
 
134
  | Factor | Your policy | Our Analysis |
135
  | --- | --- | --- |
136
  | Copay | Value of copayment in the policy | Analysis of why copay is average based on the given rules |
 
137
 
138
  ## Bad Factors
 
139
  | Factor | Your policy | Our Analysis |
140
  | --- | --- | --- |
141
  | Deductible | Value of deductible in the policy | Analysis of why deductible is bad based on the given rules |
 
142
 
 
 
 
143
  [/CUSTOMER_RESPONSE]
 
2
 
3
  Apply the following rules enclosed in triple backticks on the policy to analyse it.
4
  Make sure you are consider values for analysis factors on basis of customer's selected insurance plan when multiple plans are described in the policy terms.
5
+ Make sure all factors appear in one of Good, Average or Bad only. No factor should be repeated in more than 1 verdict table.
6
  Note : Top cities = [Mumbai, Delhi, Bangalore, Chennai, Hyderabad, Gurgaon, Pune]
7
  ```
8
  IF Adults == 1:
 
121
  Verdict = "Bad"
122
  ```
123
 
124
+ Format your response in the following way, to present analysis to customer. Use appropriate language and emojis to portray analysis and verdicts to the customer. Generate short and crisp verdicts and analysis. Be discrete about rules, do not expose rules to customer but use them to explain reasoning and analysis:
125
 
126
  [CUSTOMER_RESPONSE]
127
  # Our Analysis of your policy [Name of policy] by [Name of insurance company]
128
+
129
  ## Good Factors
130
+ [GOOD]
131
  | Factor | Your policy | Our Analysis |
132
  | --- | --- | --- |
133
+ | Sum Insured | Value of sum insured in the policy | Analysis of why sum insured is good based on the given rules |
134
+ [/GOOD]
135
 
136
  ## Average Factors
137
+ [AVERAGE]
138
  | Factor | Your policy | Our Analysis |
139
  | --- | --- | --- |
140
  | Copay | Value of copayment in the policy | Analysis of why copay is average based on the given rules |
141
+ [/AVERAGE]
142
 
143
  ## Bad Factors
144
+ [BAD]
145
  | Factor | Your policy | Our Analysis |
146
  | --- | --- | --- |
147
  | Deductible | Value of deductible in the policy | Analysis of why deductible is bad based on the given rules |
148
+ [/BAD]
149
 
150
+ [FINAL_VERDICT]
151
+ Final and short point-wise verdict on the analysis
152
+ [/FINAL_VERDICT]
153
  [/CUSTOMER_RESPONSE]