Back to projects
RAG

Foundational RAG System

Streamlit + FAISS Document Q&A

Overview

Early-stage RAG system built to explore document ingestion, embeddings, and grounded Q&A. Features multi-format document ingestion, metadata-aware chunking, and FAISS vector search with MMR reranking.

Key Capabilities

  • Multi-format document ingestion
  • Metadata-aware chunking
  • FAISS vector search with MMR reranking
  • Gemini / OpenAI embeddings
  • Streamlit-based UI with chat history

AI Concepts

Retrieval-Augmented GenerationVector SearchPrompt Grounding & Safety

Tech Stack

PythonStreamlitFAISSGoogle GeminiOpenAI