RAG Evaluation: Faithfulness, Context Recall, and Answer Relevance
RAGAs evaluation enhances AI systems' reliability by utilizing three key metrics: faithfulness, context recall, and answer relevance.
When a Vietnamese bank's AI assistant confidently quoted non-existent compliance rules, it revealed that the testing approach was fundamentally flawed. The RAGAs evaluation framework dissects RAG quality into independently measurable metrics, enabling better diagnostics of system errors. The three key metrics—faithfulness, context recall, and answer relevance—are crucial for assessing system performance.