Update src/streamlit_app.py

src/streamlit_app.py

@@ -23,8 +23,6 @@ st.markdown("""
 </style>
 """, unsafe_allow_html=True)
 
-
-
 # Header
 with st.container():
     st.markdown('<div class="header">', unsafe_allow_html=True)
@@ -36,14 +34,14 @@ with st.container():
 with st.sidebar:
     st.header("⚙️ Settings")
     
-    # API key from Hugging Face Secrets
-    try:
-        GOOGLE_API_KEY = st.secrets.get["GOOGLE_API_KEY"]
-        st.success("API Key loaded from Hugging Face Secrets!")
-    except KeyError:
-        st.error("Google API Key not found in Hugging Face Secrets. Please set 'GOOGLE_API_KEY'.")
-        GOOGLE_API_KEY = None
+    # Load API key securely from Hugging Face secrets
+    GOOGLE_API_KEY = st.secrets.get("GOOGLE_API_KEY", None)
     
+    if GOOGLE_API_KEY:
+        st.success("✅ Google API Key loaded successfully!")
+    else:
+        st.error("❌ Google API Key not found in Hugging Face Secrets. Please set 'GOOGLE_API_KEY'.")
+
     # Theme toggle
     theme = st.selectbox("Theme", ["Light", "Dark"], index=0)
     if theme == "Dark":
@@ -53,8 +51,6 @@ with st.sidebar:
 @st.cache_resource
 def initialize_vectorstore(_api_key):
     embedding = GoogleGenerativeAIEmbeddings(model="models/embedding-001", google_api_key=_api_key)
-    
-    # Extract chroma_db1.zip
     zip_path = "./chroma_db1.zip"
     extract_dir = "chroma_db2"
     if os.path.exists(zip_path):
@@ -63,54 +59,52 @@ def initialize_vectorstore(_api_key):
                 zip_ref.extractall(extract_dir)
             vectorstore = Chroma(persist_directory=extract_dir, embedding_function=embedding)
             vectorstore.persist()
-            if vectorstore._collection.count() > 0:  # Check if DB has data
+            if vectorstore._collection.count() > 0:
                 return vectorstore
             else:
-                st.error("Chroma DB from chroma_db1.zip is empty. Please provide a valid database.")
-                return None
+                st.error("Chroma DB is empty after extraction.")
         except Exception as e:
-            st.error(f"Failed to load Chroma DB from chroma_db1.zip: {str(e)}. Please check the zip file.")
-            return None
+            st.error(f"Failed to load Chroma DB: {str(e)}")
     else:
-        st.error(f"chroma_db1.zip not found at {zip_path}. Please upload the zip file.")
-        return None
+        st.error(f"`chroma_db1.zip` not found at {zip_path}")
+    return None
 
 # Initialize vectorstore and retriever
+retriever = None
 if GOOGLE_API_KEY:
-    try:
-        vectorstore = initialize_vectorstore(GOOGLE_API_KEY)
-        retriever = vectorstore.as_retriever(search_type="mmr", search_kwargs={"k": 3, "lambda_mult": 1}) if vectorstore else None
-    except Exception as e:
-        st.error(f"Error initializing vectorstore: {str(e)}")
-        retriever = None
-else:
-    retriever = None
+    vectorstore = initialize_vectorstore(GOOGLE_API_KEY)
+    if vectorstore:
+        retriever = vectorstore.as_retriever(search_type="mmr", search_kwargs={"k": 3, "lambda_mult": 1})
 
 # Prompt template
 prompt = ChatPromptTemplate.from_messages([
     ("system",
-     """
-     You are a domain-specific AI financial analyst focused on company-level performance evaluation.
-
-     Your task is to analyze and respond to user financial queries *strictly based on the provided transcript data*: {context}.
-
-     Rules:
-     1. ONLY extract facts, figures, and insights that are explicitly available in the transcript.
-     2. If data is *missing or partially available*, clearly state: "The required data is not available in the current transcript."
-     3. Do not assume or hallucinate values. Be transparent and evidence-driven.
-     4. Prioritize answers for ITC Ltd., but keep the structure reusable.
-     5. Use bullet points or structure year-wise/metric-wise data when appropriate.
-     """),
+     """You are a domain-specific AI financial analyst focused on company-level performance evaluation.
+
+Your task is to analyze and respond to user financial queries strictly based on the provided transcript data: {context}.
+
+Rules:
+1. ONLY extract facts, figures, and insights that are explicitly available in the transcript.
+2. If data is missing or partially available, clearly state: "The required data is not available in the current transcript." Then provide a generic but relevant explanation based on standard financial principles.
+3. Maintain numerical accuracy and avoid interpretation beyond data boundaries.
+4. Prioritize answers relevant to ITC Ltd., but keep response format adaptable to other firms and fiscal years.
+5. Clearly present year-wise or metric-wise insights using bullet points or structured formats if applicable.
+
+Your goals:
+- Ensure 100% fidelity to source transcript.
+- Do not assume or hallucinate missing numbers.
+- Use clear, reproducible reasoning steps (e.g., show which line items support your conclusion).
+- Output should be modular enough to scale across other companies and time periods.
+
+Respond only to this question from the user."""),
+
     ("human", "{question}")
 ])
-
 # LLM setup
+llm, parser = None, None
 if GOOGLE_API_KEY:
     llm = ChatGoogleGenerativeAI(api_key=GOOGLE_API_KEY, model="gemini-1.5-flash", temperature=1)
     parser = StrOutputParser()
-else:
-    llm = None
-    parser = None
 
 # Helper functions
 def format_docs(docs):
@@ -125,31 +119,32 @@ def retrieve_and_answer(question):
     result = (prompt | llm | parser).invoke(final_input)
     return result, docs
 
-# Streamlit UI - Query input
+# Query input form
 st.subheader("🔍 Ask a Financial Question")
 with st.form(key="query_form", clear_on_submit=True):
     query = st.text_input("Enter your question about ITC's financials:", placeholder="e.g., What was ITC's revenue in FY 2023?")
     submit_button = st.form_submit_button("Get Answer")
 
-if submit_button and query.strip() and GOOGLE_API_KEY:
-    with st.spinner("Generating answer..."):
-        try:
-            answer, source_docs = retrieve_and_answer(query)
-            st.markdown('<div class="answer-box">', unsafe_allow_html=True)
-            st.markdown("### ✅ Answer")
-            st.markdown(answer)
-            st.markdown('</div>', unsafe_allow_html=True)
-
-            with st.expander("📄 Source Documents", expanded=False):
-                if source_docs:
-                    for doc in source_docs:
-                        st.markdown(f"- **Source**: {doc.metadata.get('source', 'Unknown document')}")
-                        st.markdown(f"  **Content**: {doc.page_content}")
-                else:
-                    st.write("No source documents found.")
-        except Exception as e:
-            st.error(f"Error processing query: {str(e)}")
-elif submit_button and not query.strip():
-    st.warning("Please enter a valid question.")
-elif submit_button and not GOOGLE_API_KEY:
-    st.error("Please set the 'GOOGLE_API_KEY' in Hugging Face Secrets to proceed.")
+if submit_button:
+    if not query.strip():
+        st.warning("Please enter a valid question.")
+    elif not GOOGLE_API_KEY:
+        st.error("Google API Key not configured. Set it in Hugging Face Secrets to proceed.")
+    else:
+        with st.spinner("Generating answer..."):
+            try:
+                answer, source_docs = retrieve_and_answer(query)
+                st.markdown('<div class="answer-box">', unsafe_allow_html=True)
+                st.markdown("### ✅ Answer")
+                st.markdown(answer)
+                st.markdown('</div>', unsafe_allow_html=True)
+
+                with st.expander("📄 Source Documents", expanded=False):
+                    if source_docs:
+                        for doc in source_docs:
+                            st.markdown(f"- **Source**: {doc.metadata.get('source', 'Unknown document')}")
+                            st.markdown(f"  **Content**: {doc.page_content}")
+                    else:
+                        st.write("No source documents found.")
+            except Exception as e:
+                st.error(f"Error processing query: {str(e)}")