Skip to content
For Managed AI Platform users, see the cloud platform documentation

Workload API

Run agents, models, apps, and other services with governance from day zero: artifacts define what runs, Workloads provide the stable identity and policy, and the platform runs your containers behind that identity.

  • Get started

    Discover Workload API concepts and endpoints, then deploy your first Workload, from a project folder with dr workload up or from a published image.

  • Build artifacts

    Define a container and version it as the artifact behind a Workload.

  • Create Workloads

    Create a Workload from an artifact, give it a stable identity, and assign a draft or locked status.

  • Operate running Workloads

    Start and stop a Workload, adjust runtime settings, manage sharing, and inspect lifecycle state.

  • Update existing Workloads

    Roll out a new artifact to a running Workload as a rolling replacement, or promote a draft to production.

  • Monitor telemetry and health

    Review service statistics, OpenTelemetry traces and logs, and the console views that surface them.

  • Workload interfaces

    Deploy a project folder with dr workload up, manage artifacts and Workloads from the terminal with dr artifact and dr workload, or declare them as code with the DataRobot Pulumi provider.