Microsoft Certified: Azure AI Engineer AssociateImplement generative AI solutionsMedium
A team is deploying a generative AI model into a production environment using Azure OpenAI Service. To ensure responsible AI practices, they need to monitor the model's output for potential biases or harmful content over time. Which Azure service is best suited for continuous monitoring and logging of AI model inputs, outputs, and associated metrics?
- AAzure Data Lake Storage
- BAzure Functions
- CAzure Monitor
- DAzure DevOps
Show answer & explanationAnswer & explanation
Correct answer: C. Azure Monitor
Azure Monitor is the primary service for collecting, analyzing, and acting on telemetry data from your Azure resources, including logs and metrics from AI services. It's ideal for continuous monitoring of model inputs, outputs, and for detecting issues like bias or harmful content.
Why the other options are wrong
- A. Azure Data Lake Storage is for storing large amounts of data, not for real-time monitoring and analytics.
- B. Azure Functions are serverless compute services, useful for processing data but not for the overall monitoring infrastructure itself.
- D. Azure DevOps is for CI/CD and project management, not for operational monitoring of deployed AI models.
Azure Monitor for Responsible AI
Azure Monitor is a comprehensive monitoring solution that collects and analyzes telemetry from Azure resources, enabling continuous observation of AI model performance, inputs, and outputs to ensure responsible and ethical operation.
- Collects logs and metrics.
- Provides dashboards and alerting capabilities.
- Essential for detecting bias, drift, or harmful content over time.
Memory trick: Azure Monitor is the 'AI Watchdog' for your deployed models.