ReactFastAPILangGraphFAISSPython
AI Document Search System
FastAPI and Python-based backend delivering 90% accurate responses from 1,000+ page datasets. Features vector-based search using FAISS embeddings for advanced query processing, API interfaces supporting 100+ simultaneous requests with sub-3-second latency, MongoDB for user sessions and optimized vector storage.
Key Highlights
90% accuracy on 1,000+ page datasets
100+ concurrent requests supported
Sub-3-second response latency
Architecture
Video Demo
Gallery
Click to Expand
Click to Expand
Click to Expand
Click to Expand
Click to Expand
Click to Expand
Features
90% accurate responses from large datasets
Advanced RAG pipeline with LangGraph orchestration
Vector-based search using FAISS embeddings
100+ simultaneous requests with sub-3s latency
Smart chat history with AI-generated titles
MongoDB for user sessions
Pinecone for optimized vector storage
Tech Stack
ReactTailwind CSSFramer MotionFastAPILangChainLangGraphFAISSMongoDBPinecone
Interested in this project?
Check out the source code.