Update app.py

app.py

@@ -1,10 +1,9 @@
 import streamlit as st
-import numpy as np
+
 import requests
 from requests.auth import HTTPBasicAuth
 from transformers import pipeline
 import json
-from sentence_transformers import SentenceTransformer, util
 
 # Jira instance details
 jira_url = 'https://gen-ai-demo.atlassian.net/'
@@ -14,36 +13,18 @@ api_token = 'ATATT3xFfGF09uFJnHFeDFU0_AKBZXP98fd2ZFLxzbZfTQ4Tr17IZxD6qoPbniEvLRW
 
 # Function to get user input for Jira query using Streamlit
 def get_user_input():
-    summary_keyword = st.text_input("You:")
+    summary_keyword = st.text_input("Please give your I/P (e.g., Bug, Authentication issue):")
     max_results = st.number_input("Enter number of results to retrieve:", min_value=1, max_value=100, value=5)
     return summary_keyword, max_results
 
-# Function to find similar issues using embeddings
-def find_similar_issues(summary_keyword, jira_issues):
-    model = SentenceTransformer('all-MiniLM-L6-v2')
-
-    # Create embeddings for the input summary
-    query_embedding = model.encode(summary_keyword, convert_to_tensor=True)
-
-    if not jira_issues:
-        # If no issues from Jira, return an empty list or fallback options
-        st.write("No issues found in Jira. Using fallback dataset for embedding search.")
-        return []
-
-    # Create embeddings for the issue summaries
-    issue_summaries = [issue['fields']['summary'] for issue in jira_issues]
-    issue_embeddings = model.encode(issue_summaries, convert_to_tensor=True)
-
-    # Calculate cosine similarities between the input summary and the issue summaries
-    similarities = util.pytorch_cos_sim(query_embedding, issue_embeddings)[0]
+# Streamlit app layout
+st.title("Trouble_Ticket_Finder")
 
-    # Get top N most similar issues
-    top_results = np.argsort(similarities, axis=0)[-5:][::-1]
-    
-    return [(jira_issues[i]['key'], jira_issues[i]['fields']['summary'], similarities[i].item()) for i in top_results]
+# Get user input through Streamlit widgets
+summary_keyword, max_results = get_user_input()
 
-# Function to fetch Jira issues
-def fetch_jira_issues(summary_keyword, max_results):
+# If summary keyword is provided, make the Jira API request
+if summary_keyword:
     # Construct the JQL query based on the user input
     jql_query = f'summary ~ "{summary_keyword}"'  # Search for issues where summary contains the keyword
 
@@ -68,50 +49,16 @@ def fetch_jira_issues(summary_keyword, max_results):
     # Send the GET request to Jira API
     response = requests.get(url, headers=headers, auth=auth, params=params)
 
-    return response
-
-# Streamlit app layout
-st.title("Trouble_Ticket_Finder")
-
-# Get user input through Streamlit widgets
-summary_keyword, max_results = get_user_input()
-
-# If summary keyword is provided, make the Jira API request
-if summary_keyword:
-    # Fetch Jira issues based on summary keyword
-    response = fetch_jira_issues(summary_keyword, max_results)
-    
     # Check for successful request
     if response.status_code == 200:
         data = response.json()
         issues = data.get('issues', [])
-        
         if issues:
-            # If issues are found in Jira, display them
-            st.write("Found Jira issues matching the summary keyword:")
+            # If issues are found, display them
             for issue in issues:
                 st.write(f"**Key:** {issue['key']} - **Summary:** {issue['fields']['summary']}")
-                # Provide general elaboration of the issue (could be description or other fields)
-                st.write(f"**Description:** {issue['fields'].get('description', 'No description available')}")
         else:
             st.write("No issues found matching your summary keyword.")
-            
-            # If no issues are found in Jira, try to find similar issues based on embeddings
-            st.write("Searching for similar issues using embeddings...")
-
-            # If issues are empty, this will prevent the error and just return an empty list
-            similar_issues = find_similar_issues(summary_keyword, issues)  # This uses the Jira issues
-
-            if similar_issues:
-                st.write("Similar Issues found using embeddings:")
-                for key, summary, similarity in similar_issues:
-                    st.write(f"**Similar Issue Key:** {key}")
-                    st.write(f"**Summary:** {summary}")
-                    st.write(f"**Similarity Score:** {similarity:.4f}")
-                    # Provide general elaboration of the similar issue
-                    st.write(f"**Elaboration:** The similarity score indicates how closely this issue matches the input summary.")
-            else:
-                st.write("No similar issues found using embeddings.")
     else:
         # If the request failed, show the error details
         error_data = response.json()