Building Production-Ready RAG Systems Using Open Source Tools

In this session, we'll build a production-ready Retrieval-Augmented Generation (RAG) pipeline using an entirely open-source stack. We'll explore why naive RAG implementations fail, compare document chunking strategies, understand embeddings, vector searches with FAISS , implement hybrid retrieval, design grounded prompts to reduce Hallucinations , and evaluate retrieval quality using practical metrics.

Audience would have easily understood the RAG architecture, mathematical concepts explained in simple terms, with real world analogies for Vector search techniques.