M-034Build RAG Evaluation Suite

Description

CloudDocs Inc deployed a RAG system but has no way to measure quality. Users complain about irrelevant answers, but there's no data to guide improvements. Build an evaluation suite with test queries, ground truth answers, and automated metrics (precision, recall, MRR).

Target Function

function evaluateRAG(queries, groundTruth, ragSystem) { /* ... */ }

Intended Behavior

Evaluates a RAG system by running test queries against ground truth answers and computing precision, recall, and MRR metrics

Related Lessons