Update app.py

app.py

@@ -1,19 +1,40 @@
-# Save this as app.py
+# Save as app.py
 import streamlit as st
 import fitz  # PyMuPDF
+import faiss
 import os
+import numpy as np
+from sentence_transformers import SentenceTransformer
 import requests
 
-# Load PDF content
-def load_pdf_text(pdf_path):
+# Load SentenceTransformer model
+embedder = SentenceTransformer("all-MiniLM-L6-v2")
+
+# Load and split PDF
+def load_and_split_pdf(pdf_path):
     doc = fitz.open(pdf_path)
-    text = ""
+    chunks = []
     for page in doc:
-        text += page.get_text()
-    return text
+        text = page.get_text()
+        chunks.extend([chunk.strip() for chunk in text.split("\n") if chunk.strip()])
+    return chunks
+
+# Embed chunks and save in FAISS
+def create_faiss_index(text_chunks):
+    embeddings = embedder.encode(text_chunks)
+    index = faiss.IndexFlatL2(embeddings.shape[1])
+    index.add(np.array(embeddings))
+    return index, text_chunks, embeddings
+
+# Search in FAISS
+def search_faiss(query, index, text_chunks, k=3):
+    query_vec = embedder.encode([query])
+    D, I = index.search(query_vec, k)
+    results = [text_chunks[i] for i in I[0]]
+    return "\n".join(results)
 
-# Ask LLM using Groq API
-def ask_llm_groq(question, context, groq_api_key):
+# Call Groq API with context
+def ask_groq(query, context, groq_api_key):
     headers = {
         "Authorization": f"Bearer {groq_api_key}",
         "Content-Type": "application/json"
@@ -22,37 +43,35 @@ def ask_llm_groq(question, context, groq_api_key):
     data = {
         "model": "llama3-8b-8192",
         "messages": [
-            {"role": "system", "content": "Answer based on the provided context."},
-            {"role": "user", "content": f"Context: {context}\n\nQuestion: {question}"}
+            {"role": "system", "content": "Answer the user's question using the provided context."},
+            {"role": "user", "content": f"Context:\n{context}\n\nQuestion: {query}"}
         ]
     }
 
     response = requests.post("https://api.groq.com/openai/v1/chat/completions", headers=headers, json=data)
-    if response.status_code == 200:
-        return response.json()['choices'][0]['message']['content']
-    else:
-        return "Error: " + response.text
+    return response.json()['choices'][0]['message']['content'] if response.status_code == 200 else response.text
 
 # Streamlit UI
-st.set_page_config(page_title="Meraj Info Assistant", layout="centered")
-st.title("📄 Meraj Graphics RAG Chatbot")
+st.set_page_config(page_title="Meraj Graphics RAG Bot")
+st.title("📄 Meraj Graphics Assistant")
 
-uploaded_pdf = st.file_uploader("Upload the Info PDF", type="pdf")
+uploaded_pdf = st.file_uploader("Upload your PDF with business info", type="pdf")
 
 if uploaded_pdf:
     with open("temp.pdf", "wb") as f:
         f.write(uploaded_pdf.read())
-    context = load_pdf_text("temp.pdf")
 
     st.success("PDF uploaded successfully!")
+    chunks = load_and_split_pdf("temp.pdf")
+    index, chunk_texts, embeddings = create_faiss_index(chunks)
 
-    question = st.text_input("Ask anything about Meraj Graphics:")
-    groq_api_key = st.text_input("Enter your Groq API key", type="password")
+    query = st.text_input("Ask a question about our services:")
+    groq_key = st.text_input("Enter your Groq API key", type="password")
 
-    if st.button("Ask"):
-        if question and groq_api_key:
-            with st.spinner("Thinking..."):
-                answer = ask_llm_groq(question, context, groq_api_key)
-                st.markdown(f"**Answer:** {answer}")
+    if st.button("Get Answer"):
+        if query and groq_key:
+            context = search_faiss(query, index, chunk_texts)
+            answer = ask_groq(query, context, groq_key)
+            st.markdown(f"**Answer:** {answer}")
         else:
-            st.warning("Please enter both question and API key.")
+            st.warning("Please enter both a query and API key.")