Skip to content

Click in-app to access the full platform documentation for your version of DataRobot.

MLOps agents

Availability information

The MLOps agent feature is exclusive to DataRobot MLOps. Contact your DataRobot representative for information on enabling it.

DataRobot MLOps provides powerful tools for tracking and managing models for prediction. But what if you already have—or need to have—deployments running in your own environment? How can you monitor external models that may have intermittent or no connectivity and so may report predictions sporadically? The MLOps agents allow you to monitor and manage external models—those running outside of DataRobot MLOps. With this functionality, predictions and information from these models can be reported as part of MLOps deployments. You can use the same powerful model management tools to monitor accuracy, data drift, prediction distribution, latency, and more, regardless of where the model is running. Data provided to DataRobot MLOps provides valuable insight into the performance and health of those externally deployed models.

The MLOps agents provide:

  • The ability to manage, monitor, and get insight from all model deployments in a single system
  • API and communications constructed to ensure little or no latency when monitoring external models
  • Support for deployments that are always connected to the network and the MLOps system, as well as partially or never-connected deployments
  • The MLOps library (available in Python and Java), which can be used to monitor models written natively in those languages or to report the input and output of a model artifact in any language
  • Configuration with the Portable Prediction Server

See the associated feature considerations for additional information.

Monitoring agent

Topic Describes
Installation and configuration How to install and configure the monitoring agent.
Examples directory How to access and run monitoring agent code examples.
Use cases How to configure the monitoring agent to support various use cases.
Environment variables How to configure the monitoring agent environment variables, including those required for a containerized configuration.
Library and agent spooler configuration How to configure the MLOps library and agent to communicate through various spoolers (or buffers).
Download Scoring Code How to download model Scoring Code packaged with the monitoring agent.
Monitor external multiclass deployments How to monitor external multiclass deployments.

Management agent

Topic Describes
Installation and configuration How to install and configure the management agent.
Configure environment plugins How to use the example environment plugins as a starting point to configure the management agent for various prediction environments.
Installation for Kubernetes How to use a Helm chart to aid in the installation and configuration of the management agent and Kubernetes plugin.
Deployment status and events How to monitor the status and health of management agent deployments from the deployment inventory.
Relaunch deployments How to relaunch management agent deployments.
Force delete deployments How to delete a management agent deployment without waiting for the resolution of the deployment deletion request sent to the management agent.

Feature considerations

  • The MLOps agents run on Linux.
  • The MLOps agents don't support Windows environments.

Updated November 16, 2022
Back to top