A team has deployed a generative AI model that provides medical information. They need to collect user feedback on the accuracy of the responses and also trace each response back to the specific model version and prompt used. They are concerned about privacy and want to ensure that no personal health information is stored in the telemetry. What should they do?

Disable telemetry entirely and manually review logs for feedback

Use Azure Log Analytics to collect feedback and rely on Azure Policy to enforce data retention

Use Application Insights with custom properties for prompt and model version, and configure data masking to remove any PHI

Store feedback in Azure Blob Storage with the prompt and model version in the metadata, and enable encryption at rest

A developer is using a generative AI model to generate creative writing. The output is often too predictable and lacks variety. They want to increase the diversity of the output while still maintaining coherence. They have already tried increasing the temperature, but the output became too random. What should they do next?

Use a different model with a smaller parameter size

Increase the top-p value to 0.95 and keep the temperature moderate

Set the frequency penalty to 2.0 and the presence penalty to 0

Decrease the temperature to 0.1 and increase the max tokens

A team has deployed a generative AI model and wants to set up alerts for when the error rate exceeds a threshold. They also want to track user satisfaction. Which configuration should they implement?

Use Azure Service Health to monitor errors

Enable Azure Advisor recommendations

Use Azure Policy to enforce error thresholds

Create a metric alert in Azure Monitor and add a feedback survey to the application

What is the primary benefit of deploying a generative AI model in a container on an edge device?

Elimination of the need for model training

Reduced latency by processing data locally

Microsoft Certified:Azure AI Engineer Associate

Domain 2Objective 3

Optimize and Operationalize a Generative AI Solution AI-102 Practice Questions (Page 9)

Part of the Implement generative AI solutions domain, which accounts for 15-20% of the AI-102 exam. Microsoft does not publish an official question count, but from its 100-minute exam (~40–65 total, ~6–13 in this domain), expect 2–4 from this objective — we provide 47 practice questions to prepare you well beyond it. (estimate)

47questions here

10free pages

9concepts

15-20%of the exam

Questions 41–45

41expert · hard
A team has deployed a generative AI model that provides medical information. They need to collect user feedback on the accuracy of the responses and also trace each response back to the specific model version and prompt used. They are concerned about privacy and want to ensure that no personal health information is stored in the telemetry. What should they do?
Select an answer first
42expert · hard
A developer is using a generative AI model to generate creative writing. The output is often too predictable and lacks variety. They want to increase the diversity of the output while still maintaining coherence. They have already tried increasing the temperature, but the output became too random. What should they do next?
Select an answer first
43application · medium
A team has deployed a generative AI model and wants to set up alerts for when the error rate exceeds a threshold. They also want to track user satisfaction. Which configuration should they implement?
Select an answer first
44foundation · easy
What is the primary benefit of deploying a generative AI model in a container on an edge device?
Select an answer first
45foundation · easy
You are using the Azure OpenAI API to generate product descriptions. You want to limit the response to a maximum of 150 tokens. Which parameter should you set?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to AI-102

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Microsoft. “AI-102” is a trademark of its owner, used for identification only.

Optimize and Operationalize a Generative AI Solution AI-102 Practice Questions (Page 9)

Questions 41–45

A team has deployed a generative AI model and wants to set up alerts for when the error rate exceeds a threshold. They also want to track user satisfaction. Which configuration should they implement?

What is the primary benefit of deploying a generative AI model in a container on an edge device?

You are using the Azure OpenAI API to generate product descriptions. You want to limit the response to a maximum of 150 tokens. Which parameter should you set?

Keep studying — Implement generative AI solutions