AI-103 Certification Practice Question #65

Question

You have a Microsoft Foundry project that uses Azure AI Search to ground an agent in internal documentation.
After a recent content update, users report that the agent’s answers have become less accurate.
You need to identify whether the retrieved content is negatively influencing the model’s generated responses.
Which observability signal should you review?

Accepted Answer

The goal is to determine whether the **retrieved content is negatively influencing the model's generated responses** after a content update reduced answer accuracy. The Foundry observability signal that measures exactly the alignment between the generated response and the retrieved grounding context is the **Groundedness evaluation metric** — MS Learn defines it as measuring how consistent the response is with the retrieved context, and its troubleshooting guidance ties a low groundedness score directly to incomplete or poor retrieved context. Indexer status (A) only confirms whether ingestion succeeded, latency traces (B) address speed not correctness, and prediction drift (C) is a classic ML-model concept that doesn't capture RAG context fidelity. Suggested Answer, the embedded `✓`, and MS Learn all converge on D; confidence is strong, just below the top band given the absence of any community vote.

More AI-103 practice questions