RAG

app/main.py

@@ -12,7 +12,8 @@ from pydantic import BaseModel, ValidationError
 from app.services.message import generate_reply, send_reply
 import logging
 from datetime import datetime
-import time
+from sentence_transformers import SentenceTransformer
+from app.search.rag_pipeline import RAGSystem
 from contextlib import asynccontextmanager
 # from app.db.database import create_indexes, init_db
 from app.services.webhook_handler import verify_webhook
@@ -22,6 +23,7 @@ from app.handlers.media_handler import WhatsAppMediaHandler
 from app.services.cache import MessageCache
 from app.services.chat_manager import ChatManager
 from app.api.api_prompt import prompt_router
+from app.api.api_file import file_router
 from app.utils.load_env import ACCESS_TOKEN, WHATSAPP_API_URL, GEMINI_API
 
 # Configure logging
@@ -49,6 +51,11 @@ async def setup_message_handler():
         media_handler=media_handler,
         logger=logger
     )
+async def setup_rag_system():
+    embedding_model = SentenceTransformer('all-MiniLM-L6-v2')  # Replace with your model if different
+    rag_system = RAGSystem(embedding_model)
+
+    return rag_system
 # Initialize FastAPI app
 @asynccontextmanager
 async def lifespan(app: FastAPI):
@@ -58,7 +65,9 @@ async def lifespan(app: FastAPI):
         # await init_db()
 
         logger.info("Connected to the MongoDB database!")
+        rag_system = await setup_rag_system()
 
+        app.state.rag_system = rag_system
 
         global message_handler, webhook_handler
         message_handler = await setup_message_handler()
@@ -80,7 +89,7 @@ app.add_middleware(SlowAPIMiddleware)
 
 # app.include_router(users.router, prefix="/users", tags=["Users"])
 app.include_router(prompt_router, prefix="/prompts", tags=["Prompts"])
-
+app.include_router(file_router, prefix="/file_load", tags=["File Load"])
 # Prometheus metrics
 webhook_requests = Counter('webhook_requests_total', 'Total webhook requests')
 webhook_processing_time = Histogram('webhook_processing_seconds', 'Time spent processing webhook')