app.py

import os
import gradio as gr
import faiss
import numpy as np
from groq import Groq
from pypdf import PdfReader
from sentence_transformers import SentenceTransformer

# -----------------------------
# Load Groq Client
# -----------------------------
GROQ_API_KEY = os.getenv("GROQ_API_KEY")

if not GROQ_API_KEY:
    raise ValueError("GROQ_API_KEY not found in environment variables")

client = Groq(api_key=GROQ_API_KEY)

# -----------------------------
# Load Embedding Model (Open Source)
# -----------------------------
embedding_model = SentenceTransformer("all-MiniLM-L6-v2")

# Global storage
documents = []
index = None


# -----------------------------
# PDF Processing
# -----------------------------
def process_pdf(file):
    global documents, index
    
    reader = PdfReader(file.name)
    text = ""
    
    for page in reader.pages:
        text += page.extract_text()
    
    # Chunking
    chunk_size = 500
    chunks = [text[i:i+chunk_size] for i in range(0, len(text), chunk_size)]
    
    documents = chunks
    
    # Create embeddings
    embeddings = embedding_model.encode(chunks)
    embeddings = np.array(embeddings).astype("float32")
    
    # Create FAISS index
    dimension = embeddings.shape[1]
    index = faiss.IndexFlatL2(dimension)
    index.add(embeddings)
    
    return "✅ PDF processed and indexed successfully!"


# -----------------------------
# Question Answering
# -----------------------------
def answer_question(question):
    global documents, index
    
    if index is None:
        return "Please upload and process a PDF first."
    
    # Embed question
    question_embedding = embedding_model.encode([question])
    question_embedding = np.array(question_embedding).astype("float32")
    
    # Search in FAISS
    k = 3
    distances, indices = index.search(question_embedding, k)
    
    context = ""
    for idx in indices[0]:
        context += documents[idx] + "\n\n"
    
    prompt = f"""
    Answer the question based only on the context below.

    Context:
    {context}

    Question:
    {question}

    Answer:
    """
    
    # Send to Groq LLM
    chat_completion = client.chat.completions.create(
        messages=[{"role": "user", "content": prompt}],
        model="llama-3.3-70b-versatile",
    )
    
    return chat_completion.choices[0].message.content


# -----------------------------
# Gradio UI
# -----------------------------
with gr.Blocks() as app:
    gr.Markdown("# 📚 RAG PDF Chatbot (Groq + FAISS)")
    
    file_input = gr.File(label="Upload PDF")
    upload_button = gr.Button("Process PDF")
    upload_output = gr.Textbox()
    
    question_input = gr.Textbox(label="Ask a Question")
    answer_button = gr.Button("Get Answer")
    answer_output = gr.Textbox()
    
    upload_button.click(process_pdf, inputs=file_input, outputs=upload_output)
    answer_button.click(answer_question, inputs=question_input, outputs=answer_output)

app.launch()