Document Q&A System

LangChain FAISS Streamlit LLM NLP

A production-ready Retrieval-Augmented Generation (RAG) system.

What it does

Users upload documents (PDF, DOCX, TXT) and ask natural language questions. The system retrieves the most relevant chunks and answers with source citations.

Capabilities

Document ingestion and chunking
Semantic search via FAISS vector store
LLM-powered Q&A with citations
Multi-document support
Streamlit interface

Architecture

Upload → Parse → Chunk → Embed → FAISS Index → Query → LLM → Answer + Sources

View on GitHub