AI-300: Operationalizing Machine Learning and Generative AI SolutionsDemo

Question

A team deploys a generative AI application that uses a model deployed in Microsoft Foundry. The application must support latency monitoring under production load.

You need to enable performance observability.

Which three actions should you carry out, in sequence?

Drag & Drop

Set up the model endpoint in Foundry.

Enable logging.

Apply a term blocklist.

Generate token embeddings.

Configure monitoring metrics.

Explanation

A model deployment endpoint is required before deployment-level monitoring can be configured. Monitoring metrics provide operational measures such as latency and request counts, while logging retains request and response telemetry, including latency, for analysis. Term blocklists are content-safety controls, and embeddings support semantic retrieval rather than performance monitoring.

Learn more

Monitor model deployments in Microsoft Foundry Models

AI-300: Operationalizing Machine Learning and Generative AI SolutionsDemo

AstroTutor

Notes

Readiness

QuestionQ14

Community Discussion