Update app.py

app.py

@@ -1,105 +1,142 @@
 import streamlit as st
 import os
+from langchain_openai import ChatOpenAI, OpenAIEmbeddings
+from langchain_community.document_loaders import TextLoader, PyPDFLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain.vectorstores import FAISS
+from langchain.prompts import ChatPromptTemplate
+from langchain_core.output_parsers import StrOutputParser
+import tempfile
 import time
-from langchain_openai import ChatOpenAI
-from openai import RateLimitError
 
-# YOUR API KEY (with rate limit handling)
+# YOUR API KEY
 os.environ["OPENAI_API_KEY"] = "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"
 
-KNOWLEDGE = """
-SR University is located in Warangal, Telangana, India. 
-Computer Science program focuses on AI/ML, DSA, Java/Python, AWS/Azure, software engineering.
-
-B.Tech student preparing for AI/ML internships. Skills: LeetCode, AI projects (robotic arms, drones), 
-cloud certifications, competitive programming.
-
-Internship tips:
-1. 300+ LeetCode (Easy:100, Medium:150, Hard:50)
-2. 3 portfolio projects: RAG chatbot, object detection, RL agent
-3. Apply startups: AngelList, Y Combinator
-4. Practice system design, behavioral interviews
-5. Target: Google, Microsoft, Hyderabad/Bangalore startups
-"""
-
-def get_chat_response(query, max_retries=3):
-    """Handle rate limits with retry + fallback"""
-    llm = ChatOpenAI(model="gpt-4o-mini", temperature=0)
-    
-    for attempt in range(max_retries):
-        try:
-            prompt = f"""Use ONLY this context to answer: 
-
-{KNOWLEDGE}
-
-Question: {query}
-
-Answer:"""
-            
-            response = llm.invoke(prompt).content
-            return response
-            
-        except RateLimitError:
-            if attempt < max_retries - 1:
-                st.warning(f"Rate limit hit. Retrying in 10s... (Attempt {attempt + 1}/{max_retries})")
-                time.sleep(10)
-                continue
-            else:
-                # FALLBACK: Local rule-based response
-                return get_fallback_response(query)
+@st.cache_resource
+def load_vectorstore(_file_path):
+    """Load or create vectorstore from uploaded file"""
+    if os.path.exists("faiss_index"):
+        embeddings = OpenAIEmbeddings()
+        vectorstore = FAISS.load_local("faiss_index", embeddings, allow_dangerous_deserialization=True)
+        return vectorstore
     
-    return "Sorry, service temporarily unavailable. Try again later."
-
-def get_fallback_response(query):
-    """FREE fallback - no API needed"""
-    query_lower = query.lower()
+    # Process uploaded file
+    file_ext = _file_path.split('.')[-1].lower()
+    if file_ext == 'pdf':
+        loader = PyPDFLoader(_file_path)
+    else:
+        loader = TextLoader(_file_path)
     
-    if "university" in query_lower or "sr" in query_lower:
-        return "SR University is in Warangal, Telangana. Excellent CS program with AI/ML, DSA, cloud computing focus."
+    docs = loader.load()
+    text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
+    splits = text_splitter.split_documents(docs)
     
-    elif "internship" in query_lower or "job" in query_lower:
-        return """AI/ML Internship Tips:
-1. Solve 300+ LeetCode problems
-2. Build RAG chatbot (this project!)
-3. Apply via AngelList, LinkedIn
-4. Target Hyderabad/Bangalore startups"""
+    embeddings = OpenAIEmbeddings()
+    vectorstore = FAISS.from_documents(splits, embeddings)
+    vectorstore.save_local("faiss_index")
     
-    elif "project" in query_lower:
-        return "Portfolio projects: 1) RAG Chatbot (live now), 2) Object Detection (robotic arm), 3) RL Drone Agent"
+    st.sidebar.success("✅ Vector index created from your document!")
+    return vectorstore
+
+def get_rag_chain(vectorstore):
+    """Create RAG chain"""
+    retriever = vectorstore.as_retriever(search_kwargs={"k": 4})
+    llm = ChatOpenAI(model="gpt-4o-mini", temperature=0)
     
-    elif "leetcode" in query_lower:
-        return "LeetCode: Easy(100) + Medium(150) + Hard(50) = 300 problems total"
+    prompt = ChatPromptTemplate.from_template(
+        """You are a helpful assistant. Answer using ONLY the following context:
+        
+{context}
+        
+Question: {question}
+        
+Answer concisely and accurately. If answer not in context, say "I don't have that information.""""
+    )
     
-    else:
-        return "Ask about SR University, internships, LeetCode, or AI projects! 💡"
+    chain = (
+        {"context": retriever, "question": lambda x: x} 
+        | prompt 
+        | llm 
+        | StrOutputParser()
+    )
+    return chain
 
-st.title("🧠 RAG Chatbot")
-st.info("💡 Works even if OpenAI quota exhausted!")
+# === MAIN UI ===
+st.title("🧠 Dynamic RAG Chatbot")
+st.markdown("**Upload ANY document → Ask ANY question → Get accurate answers!**")
 
-if "messages" not in st.session_state:
-    st.session_state.messages = []
+# File upload
+uploaded_file = st.file_uploader("📤 Upload PDF or TXT", type=['pdf', 'txt'])
 
-# Chat history
-for message in st.session_state.messages:
-    with st.chat_message(message["role"]):
-        st.markdown(message["content"])
+if uploaded_file is not None:
+    # Save uploaded file
+    with tempfile.NamedTemporaryFile(delete=False, suffix=f".{uploaded_file.name.split('.')[-1]}") as tmp_file:
+        tmp_file.write(uploaded_file.getvalue())
+        file_path = tmp_file.name
+    
+    st.success(f"✅ Loaded: {uploaded_file.name}")
+    
+    # Load vectorstore
+    with st.spinner("🔄 Indexing your document..."):
+        try:
+            vectorstore = load_vectorstore(file_path)
+            chain = get_rag_chain(vectorstore)
+            st.session_state.chain = chain
+            st.session_state.ready = True
+        except Exception as e:
+            st.error(f"Error: {str(e)}")
+            st.session_state.ready = False
 
-# Chat input
-if query := st.chat_input("Ask about university, internships, projects..."):
-    st.session_state.messages.append({"role": "user", "content": query})
-    with st.chat_message("user"):
-        st.markdown(query)
+# Chat interface (only if ready)
+if 'ready' in st.session_state and st.session_state.ready:
+    st.success("🚀 Chatbot ready! Ask questions about your document.")
     
-    with st.chat_message("assistant"):
-        with st.spinner('Thinking...'):
-            response = get_chat_response(query)
-            st.markdown(response)
+    if "messages" not in st.session_state:
+        st.session_state.messages = []
     
-    st.session_state.messages.append({"role": "assistant", "content": response})
+    # Chat history
+    for message in st.session_state.messages:
+        with st.chat_message(message["role"]):
+            st.markdown(message["content"])
+    
+    # Chat input
+    if query := st.chat_input("💬 Ask ANY question about your document..."):
+        st.session_state.messages.append({"role": "user", "content": query})
+        with st.chat_message("user"):
+            st.markdown(query)
+        
+        with st.chat_message("assistant"):
+            with st.spinner("Searching document..."):
+                try:
+                    response = st.session_state.chain.invoke(query)
+                    st.markdown(response)
+                except Exception as e:
+                    st.error(f"Error: {str(e)}")
+                    st.markdown("Try re-uploading your document.")
+        
+        st.session_state.messages.append({"role": "assistant", "content": response})
+
+    # Clear chat
+    if st.button("🗑️ Clear Chat"):
+        st.session_state.messages = []
+        st.rerun()
 
-# Clear chat
-if st.sidebar.button("🗑️ Clear Chat"):
-    st.session_state.messages = []
-    st.rerun()
+else:
+    st.info("👆 **Upload a PDF or TXT file to start chatting!**")
+    st.markdown("""
+    **Works with:**
+    • Resumes/CVs (.pdf, .txt)
+    • Research papers
+    • Notes/lecture slides
+    • Company docs
+    • ANY text content!
+    """)
 
-st.sidebar.success("✅ Works with/without OpenAI quota!")
+# Sidebar
+with st.sidebar:
+    st.markdown("### 🎯 Features")
+    st.markdown("- Dynamic document upload")
+    st.markdown("- True RAG (vector search)")
+    st.markdown("- Handles PDF/TXT")
+    st.markdown("- Production UI")
+    st.markdown("- Portfolio ready!")