RAG evaluation template
The evaluation structure we use on every retrieval-augmented system. Copy it, fill in your numbers, and you will know whether the system is trustworthy before users do.
Golden set
Retrieval metrics
Answer metrics
Operational metrics
Release gate
Want help with the gaps?
This checklist comes from our Generative AI Engineering work. Send us the unticked items.