← CurriculumWeek 05 / 10 · Systems
05
RAG: Retrieval-Augmented Generation
Understand how AI accesses information it wasn't trained on — and learn the discipline of choosing RAG vs. simpler patterns.
Topics covered
- 01What RAG is and what problem it actually solves
- 02Vector embeddings explained without math
- 03Vector databases: Pinecone, Chroma, Supabase Vector — what to pick when
- 04When to RAG vs. when to just paste context
- 05Chunking, embedding, retrieval, re-ranking — the four-stage pipeline
- 06RAG failure modes: bad chunks, irrelevant retrieval, stale data
- 07Building and evaluating a real pipeline end-to-end
Build challenge
Deliverable
Working RAG bot over a real Outpost knowledge base — must answer 5 evaluation queries correctly with cited sources.