DocuQuery-AI: Multi-PDF RAG Assistant

DocuQuery-AI A high-performance RAG-based AI chatbot built with LangChain, FAISS, and Google Gemini 2.5 Flash to chat with multiple PDF documents in real-time

---

Key Features

• Multi-Document Analysis: Upload and query multiple PDF files simultaneously.
• Contextual Accuracy: Uses semantic search to ensure answers are grounded in your data.
• Scalable Vector Search: Powered by FAISS (Facebook AI Similarity Search) for millisecond retrieval.
• Modern UI: A clean, intuitive interface built with Streamlit.

Technical Architecture

1. Ingestion: PDFs are parsed using PyPDF2.
2. Chunking: Text is split into segments using RecursiveCharacterTextSplitter.
3. Embeddings: Text chunks are converted into vectors using models/gemini-embedding-001.
4. Vector Store: Store indexed in a local FAISS database.
5. LLM: Gemini 2.5 Flash generates detailed responses based on retrieved context.

---