Engineer: Designing and Implementing an Azure [AI-100] Flashcards

You want to implement Responsible AI principles in your solution by detecting bias in model predictions across demographic groups. Which Azure tool supports this?

Answer: Fairlearn integrated with Azure Machine Learning

Fairlearn, integrated into Azure Machine Learning, provides metrics and visualizations to assess and mitigate model unfairness across demographic groups.

Your team needs to share and version ML datasets across multiple Azure ML workspaces. What is the recommended approach?

Answer: Register datasets in each workspace using Azure ML registered datasets pointing to the same storage

Registering datasets in Azure ML workspaces with versioning lets teams track, share, and reproduce data assets while the underlying data lives in shared storage.

You are configuring an Azure Cognitive Search skillset and need to call a custom Python function to enrich documents. What feature enables this?

Answer: Custom Web API skill pointing to an Azure Function endpoint

The Custom Web API skill in Azure Cognitive Search lets you call any HTTP endpoint, such as an Azure Function, to apply custom enrichment logic.

When designing an Azure AI solution, you want to ensure that the system can explain individual predictions to end users. Which technique is most applicable?

Answer: Model interpretability using SHAP values in Azure ML

SHAP (SHapley Additive exPlanations) values, available through Azure ML's interpretability library, explain the contribution of each feature to individual predictions.

Your Azure Bot sends proactive messages to users in Microsoft Teams. Which Bot Framework feature must be stored and reused to send proactive messages?

Answer: The conversation reference obtained from a previous turn

A conversation reference captured during an active turn contains the service URL and conversation ID needed to initiate proactive messages later.

You need to extract structured data from handwritten forms with variable layouts that are not covered by prebuilt Form Recognizer models. What should you use?

Answer: A custom Form Recognizer model trained with labeled sample forms

Custom Form Recognizer models learn field locations from labeled training documents, enabling accurate extraction from variable-layout forms.

Your organization's AI policy requires that all Cognitive Services calls are logged for auditing. Which Azure service should you integrate to capture and store these logs?

Answer: Azure Monitor Diagnostic Settings routed to a Log Analytics workspace

Enabling Diagnostic Settings on the Cognitive Services resource and routing logs to a Log Analytics workspace captures all request and response telemetry for auditing.