rag_integration.py

"""
RAG Integration Example

This file demonstrates how to integrate the RAG system with your existing app.py
without modifying the original code. This shows a separate example that you can
use as a reference when you're ready to integrate RAG functionality.

Usage:
1. Run train_rag.py to build your knowledge base
2. Use this integration example as a template for modifying your app.py
"""

from dotenv import load_dotenv
from openai import OpenAI
import json
import os
import requests
import gradio as gr
from rag_utils import RAGSystem
import logging

# Set up logging
logging.basicConfig(level=logging.INFO, 
                    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
logger = logging.getLogger(__name__)

load_dotenv(override=True)

# Initialize the RAG system with error handling
rag = RAGSystem()

# Try to load the pre-trained index
try:
    rag_index_path = "me/rag_index"
    if os.path.exists(rag_index_path):
        logger.info(f"Loading RAG index from {rag_index_path}")
        rag.load_index(rag_index_path)
        logger.info(f"Loaded {len(rag.chunks)} chunks from index")
    else:
        logger.warning(f"RAG index not found at {rag_index_path}. Run train_rag.py first.")
        # Create a flag to indicate missing index
        rag.index_loaded = False
except Exception as e:
    logger.error(f"Error loading RAG index: {e}")
    # Create a flag to indicate error
    rag.index_loaded = False
else:
    # Mark index as successfully loaded
    rag.index_loaded = True


class MeWithRAG:
    """Example class showing how to integrate RAG with your existing Me class"""
    
    def __init__(self):
        self.openai = OpenAI(api_key=os.getenv("GOOGLE_API_KEY"), 
                           base_url="https://generativelanguage.googleapis.com/v1beta/openai/")
        self.name = "Sagarnil Das"
        # No need to load the entire PDF and summary here!
        # Just use a brief intro for the system prompt
        self.intro = "I'm a software engineer and data scientist with expertise in AI and machine learning."
        
    def system_prompt(self, query=None):
        """Dynamic system prompt that includes RAG context if a query is provided"""
        system_prompt = f"You are acting as {self.name}. You are answering questions on {self.name}'s website, \
particularly questions related to {self.name}'s career, background, skills and experience. \
Your responsibility is to represent {self.name} for interactions on the website as faithfully as possible."
        
        # If we have a query and RAG index is loaded, get relevant context
        if query and getattr(rag, 'index_loaded', False):
            try:
                context = rag.get_context_for_query(query, top_k=3)
                system_prompt += f"\n\n## Relevant Background Information:\n{context}\n\n"
            except Exception as e:
                logger.error(f"Error retrieving context: {e}")
                system_prompt += f"\n\n## Brief Introduction:\n{self.intro}\n\n"
        else:
            # Just use a brief intro for the initial prompt
            system_prompt += f"\n\n## Brief Introduction:\n{self.intro}\n\n"
            
        system_prompt += f"With this context, please chat with the user, always staying in character as {self.name}."
        return system_prompt
    
    def chat(self, message, history):
        """Chat function that uses RAG to retrieve relevant context"""
        try:
            # First message includes base system prompt
            messages = [{"role": "system", "content": self.system_prompt()}]
            
            # Add conversation history
            if isinstance(history, list) and all(isinstance(h, dict) for h in history):
                messages.extend(history)
            else:
                for user_msg, assistant_msg in history:
                    messages.append({"role": "user", "content": user_msg})
                    messages.append({"role": "assistant", "content": assistant_msg})
            
            # Add the user's message
            messages.append({"role": "user", "content": message})
            
            # Now get a dynamic system prompt with RAG context for this specific query
            # Replace the initial system message with one that includes relevant context
            try:
                messages[0] = {"role": "system", "content": self.system_prompt(query=message)}
            except Exception as e:
                logger.error(f"Error generating system prompt: {e}")
                # Keep original system prompt if there's an error
            
            # Call the LLM
            response = self.openai.chat.completions.create(
                model="gemini-2.0-flash",
                messages=messages
            )
            
            return response.choices[0].message.content
        except Exception as e:
            logger.error(f"Error in chat method: {e}")
            return f"I apologize, but I encountered an error while processing your request. Please try again later."


# Example of how to use this
if __name__ == "__main__":
    # This is just a demonstration - not meant to be run directly
    me_rag = MeWithRAG()
    gr.ChatInterface(me_rag.chat, type="messages").launch()
    
    # When integrating with your actual app.py, you could:
    # 1. Update the Me class to use RAG-based context retrieval
    # 2. Keep all your existing tool-calling functionality
    # 3. Use the RAG system to provide context-aware responses