Skip to main content

GCP Ai Platform Model Deployment Monitoring Job

A Model Deployment Monitoring Job in Google Cloud Vertex AI (formerly AI Platform) continuously evaluates the quality and data drift of models that have been deployed to an Endpoint. The job collects real-time prediction data, compares it with a baseline, triggers alerts when thresholds are breached and produces monitoring reports that help you spot issues before they affect production.
Official documentation: https://cloud.google.com/vertex-ai/docs/reference/rest/v1/projects.locations.modelDeploymentMonitoringJobs

Supported Methods​

  • GET: Get a gcp-ai-platform-model-deployment-monitoring-job by its "locations|modelDeploymentMonitoringJobs"
  • LIST
  • SEARCH: Search Model Deployment Monitoring Jobs within a location. Use the location name e.g., 'us-central1'

gcp-monitoring-notification-channel​

A Monitoring Job can include an alertConfig that references one or more Cloud Monitoring Notification Channels. Overmind surfaces this link so you can trace which e-mail, SMS or Pub/Sub targets will receive alerts raised by the job.

gcp-cloud-kms-crypto-key​

If you configure encryptionSpec on the Monitoring Job, the results and any generated datasets will be encrypted with a customer-managed Cloud KMS Crypto Key. Overmind records this dependency so you can see which key protects the monitoring artefacts.

gcp-ai-platform-endpoint​

A Monitoring Job is always attached to one or more Vertex AI Endpoints. The job pulls prediction traffic from these Endpoints to evaluate drift and performance. Overmind shows this link to reveal exactly which live services are under surveillance.

gcp-ai-platform-model​

Each Deployed Model referenced by the Endpoint—and therefore by the Monitoring Job—originates from a Vertex AI Model resource. Linking to the Model allows you to examine the training artefact, version history and other metadata connected to what the job is monitoring.