Add an Azure trace connection

  • Release version: Zurich
  • Updated June 30, 2026
  • 3 minutes to read

    • Monitor AI agents running on Microsoft Azure by adding an Azure trace connection. AI Control Tower collects trace data through your Azure credentials and a MID Server, without requiring SDK instrumentation.

    Before you begin

    Confirm the following:

    • An active MID Server is installed and configured in your ServiceNow instance. See MID Server installation.
    • Credentials for each Azure source system you plan to configure are available.
      • Each credential must be created in Azure for the source system it applies to. For details, see the Azure Trace Collector Credentials Configuration [KB3144350] article in Now Support.
      • After each credential is created in Azure, work with your instance administrator to store it as a record in All > Connections & Credentials > Credentials.

    Role required: sn_ai_governance.ai_steward

    Procedure

    1. Navigate to All > AI Control Tower > Home > Settings > Integrations > Traces.
    2. On the Available sub-tab, select Azure.
    3. Enter a descriptive name for the connection.
      The name distinguishes this connection from others you create. For instance, you may choose a name that identifies the account, project, or environment.
    4. Select the Azure source systems to integrate with.
      • Classic Foundry — collects traces from Azure AI Foundry (classic).
      • New Foundry — collects traces from the updated Azure AI Foundry experience.
      • Application Insights — collects traces from Azure Monitor Application Insights.
    5. Select Next.
    6. Fill in the credentials for each source system you selected.
      If you selected multiple source systems, provide the credentials for the first system and then select Next to continue to the credentials page for the next source system.
      Source systemSteps
      Classic Foundry
      1. Select the name of the credential in the Azure AI Services Credential Alias field.
      2. Enter the interval, in minutes, at which the MID Server polls for new trace data in the Collection frequency (minutes) field.

        The default is 30. Set a lower value to return results sooner or set a higher value to reduce overhead for lower-volume systems.

      3. Select the name of the credential in the Azure Machine Learning Credential Alias field.
      4. Select the MID Server that runs trace collection.

        The MID Server must be active and validated. Select Go to Mid server installation to install or configure one.

      5. Select Active to begin collecting traces when you save. Clear this option to save the connection without starting collection. You can activate the connection later from its record.
      New Foundry
      1. Select the name of the OAuth 2.0 credential in the Azure Machine Learning Credential field.
      2. Select the name of the API key credential in the Application Insights Credential field.
      3. Enter the interval, in minutes, at which the MID Server polls for new trace data in the Collection frequency (minutes) field.

        The default is 30. Set a lower value to return results sooner or set a higher value to reduce overhead for lower-volume systems.

      4. Enter the Application Insights Application ID.
      5. Select the MID Server that runs trace collection.

        The MID Server must be active and validated. Select Go to Mid server installation to install or configure one.

      6. Select Active to begin collecting traces when you save. Clear it to save the connection without starting collection. You can activate the connection later from its record.
      Application Insights
      1. Select the name of the OAuth 2.0 credential in the Azure App Insights Credential field.
      2. Enter the Application Insights Resource ID.
      3. Enter the interval, in minutes, at which the MID Server polls for new trace data in the Collection frequency (minutes) field.

        The default is 30. Set a lower value to return results sooner or set a higher value to reduce overhead for lower-volume systems.

      4. Select the name of the OAuth 2.0 credential in the Azure Machine Learning Credential field.
      5. Select the MID Server that runs trace collection.

        The MID Server must be active and validated. Select Go to Mid server installation to install or configure one.

      6. Select Active to begin collecting traces when you save. Clear it to save the connection without starting collection. You can activate the connection later from its record.
    7. Select Save.

    Result

    One or more trace connections appear on the Established sub-tab. If the connection is active, AI Control Tower begins collecting trace data after the first polling interval.

    What to do next

    Choose which metrics to include in evaluation scoring. See Activate evaluation scoring for external AI systems.