Azure AI Foundry metrics integration

Use the Azure AI Foundry metrics integration to collect supported Azure Monitor usage metrics and load them into Companion Metrics in Yarken.

Benefits

The integration helps you:

• Automate the collection of Azure AI usage metrics.

• Monitor usage across multiple Azure AI resources.

• Analyze token consumption, request volumes, response performance, and cache efficiency.

• Use the collected metrics in Analytics and Custom Dashboards.

• Compare usage across configured Azure resources.

What the integration collects

The integration retrieves supported Azure Monitor metrics for the Azure AI resources configured for your environment.

Before you begin

Before you authorize the integration, make sure the following information is available:

• Microsoft Entra ID Tenant ID

• Client ID

• Client secret

• Azure region

• The Azure AI resource IDs that Yarken should monitor

• A service principal with Monitoring Reader or equivalent access to the configured resources

How the integration works

The integration follows this workflow:

1. Yarken configures the Azure AI Foundry Advanced Integration.

2. The integration authenticates using a Microsoft Entra ID service principal.

3. A scheduled job runs once every week.

4. During each execution, Yarken retrieves Azure Monitor metrics for the configured resources.

5. Usage data (Daily metric values) is ingested directly into the Companion Metrics table.

6. The data becomes available for dashboards, Analytics, and FinOps reporting.

Authorize the Azure AI Foundry connection

1. Go to Admin > Pipelines > Connections.

2. Select the ADVANCED tab.

3. Open the Azure AI Foundry connection deployed for your environment.

4. Enter the required connection details.

5. Select Connect.

6. Confirm that the connection status changes to Connected.

For the general connection workflow, see Authorize Advanced Connections.

Run and monitor the Advanced pipeline

After the connection shows Connected:

1. Go to Admin > Pipelines > Pipelines.

2. Select the ADVANCED tab.

3. Locate the Azure AI Foundry metrics pipeline.

4. Select Run when you need an immediate execution, or keep recurring execution active as configured for your environment.

5. Open the pipeline to review Last run, Last status, Succeeded, Failed, and run history.

See Manage Advanced pipelines for the common pipeline controls.

Schedule and data collection

The integration runs automatically once each week.

Each weekly execution retrieves daily Azure Monitor metric values for the configured collection period. Yarken stores each day's values as individual Companion Metric records so you can analyze historical trends at a daily level.

Automatic metric creation

During the first successful execution, Yarken automatically creates the required Metric and Companion Metric definitions.

You do not need to manually create these definitions before running the integration. Later executions reuse the existing definitions and import the latest metric values.

After the definitions have been created, subsequent executions reuse the existing definitions and import only the latest Azure AI usage metrics.

Monitor multiple Azure AI resources

The integration can collect metrics for multiple configured Azure AI resources.

Yarken processes each configured resource independently and retains the source resource identification with the imported metric data. This helps you distinguish usage from different Azure resources when validating or reporting on the results.

Resource identification

Each imported metric includes the Azure Resource ID of the Azure AI Foundry resource from which it was collected.

The Resource ID is stored in the configured Companion Metrics field for the integration. This allows Yarken to distinguish metrics collected from multiple Azure AI resources and provides traceability when building reports or troubleshooting imported data.

When multiple Azure AI resources are monitored, each metric remains associated with its originating resource, preventing metrics from different resources from being combined.

Where the data appears

After a successful run, the imported metrics are available in:

• Companion Metrics

• Analytics, where the Companion Metrics data is exposed for analysis

• Custom Dashboards, when you use the available metrics in dashboard visualizations

Validate the integration

After the first successful run, confirm that:

• The Azure AI Foundry Advanced Connection shows Connected.

• The Advanced pipeline completed successfully.

• The expected daily metric records are available in Companion Metrics.

• The metric dates match the expected collection period.

• Metrics from multiple configured resources remain distinguishable.

Credential rotation

If you rotate or replace the Microsoft Entra client secret, update the Azure AI Foundry Advanced Connection with the replacement secret.

After the update, confirm that the connection returns to Connected, then run the Advanced pipeline to validate the integration.

Do not expose the client secret in screenshots, documentation, email, chat, or support tickets.

Availability

The Azure AI Foundry metrics integration is available on request. Contact your Yarken Customer Success or delivery representative if the integration has not been enabled for your environment.

Current limitations

The current integration:

• Collects metrics on a weekly schedule.

• Retrieves aggregated Azure Monitor metrics rather than raw event-level usage data.

Troubleshoot the integration

For general Advanced Integration issues, see Troubleshoot Advanced Integrations.

Related Microsoft documentation

• https://learn.microsoft.com/en-us/azure/ai-foundry/

• https://learn.microsoft.com/en-us/azure/azure-monitor/essentials/metrics-supported

• https://learn.microsoft.com/en-us/entra/identity-platform/app-objects-and-service-principals

Related Yarken content

• Advanced integrations

• Authorize Advanced Connections

• Manage Advanced pipelines