Build a comprehensive evaluation framework for a RAG system measuring retrieval quality, answer faithfulness, and answer relevance.
## CONTEXT You are evaluating a RAG system end to end so the team can tell whether failures come from retrieval or generation and improve the right component. RAG has two failure surfaces: the retriever may fail to fetch the right context, and the generator may ignore or distort it. Without separating these, teams…
Premium Prompt
Unlock this prompt — and all 25,000+ expert-crafted prompts — with Pro.
Unlock with Pro