Question 130
At an enterprise maintaining a deployed AI service, an AI project leader faces this situation: A production inference API has intermittent failures even though model accuracy reports remain stable. Governance and technical measures need accountable ongoing operation. Which use case most closely matches this proposed approach?
Explanation
Correct Answer
In an enterprise maintaining a deployed AI service, this is the matching use case: Use technical health monitoring to operate AI as a dependable service. The reason is: Model correctness and serving availability are separate dimensions of operational performance. In the reported scenario, a production inference API has intermittent failures even though model accuracy reports remain stable. The appropriate action is: Monitor request success rate, tail latency, throughput and saturation against service objectives.
Incorrect Answers
Project closeout should reconcile delivered functionality and realized value with the original business case. The appropriate correction for the reported case is: Monitor request success rate, tail latency, throughput and saturation against service objectives.
A service handoff needs explicit operational expectations and escalation arrangements. Fixing that different mechanism would not address the current root cause: model correctness and serving availability are separate dimensions of operational performance.
External dependencies and renewal obligations must be transferred with the service, not left with a departing individual. Its verification is distinct from the reported condition: A production inference API has intermittent failures even though model accuracy reports remain stable. Governance and technical measures need accountable ongoing operation.