Hands-on: Build a RAG app with LangChain and pgvector

A full-day, laptop-open workshop. By the end you will have a working retrieval-augmented app deployed to a free tier, with evals you can trust.

What we cover

1. Chunking and embedding strategies that actually matter
2. pgvector indexes (HNSW vs IVFFlat) and when to use each
3. Hybrid search with BM25
4. Building an eval set and measuring regressions
5. Deploying with a ₹0 bill

Prerequisites

• Comfortable with Python or TypeScript
• A laptop with Docker installed

Lunch and coffee included. Limited to 40 seats so everyone gets help.