get_project
Get Project
Retrieve detailed information about a specific DataRobot project including its configuration, target, stage, partition settings, and advanced options.
get_project
Retrieve detailed information about a specific DataRobot project including its configuration, target, stage, partition settings, and advanced options.
list_datasets
List datasets in the DataRobot AI Catalog. Returns metadata including name, size, row/column counts, and processing state.
get_deployment_monitoring
Retrieve monitoring data for a deployed model including service health statistics, accuracy metrics, data drift, and target drift. Choose which monitoring aspects to include.
make_predictions
Make real-time predictions using a deployed model. Pass an array of data rows as key-value objects. Optionally include prediction explanations to understand why the model makes each prediction.
start_autopilot
Set the target variable and start Autopilot on an existing project. Autopilot automatically selects and trains the best predictive models for the specified target feature. Supports configuration of mode, metric, and advanced options.
get_deployment
Retrieve detailed information about a specific model deployment including its health status, model information, capabilities, and configuration.
list_deployments
List all model deployments with their health status, importance, and prediction usage. Useful for monitoring deployed models across the organization.
create_deployment
Deploy a trained model to production. Supports deploying from a learning model (by modelId) or from a registered model package (by modelPackageId). The deployment will be accessible for real-time or batch predictions.
manage_dataset
Manage datasets in the AI Catalog. Create a new dataset from a URL, update an existing dataset's name, or delete a dataset.
list_models
List trained models in a DataRobot project, including the recommended model. Returns model types, training details, and performance metrics from the leaderboard.
manage_deployment
Manage an existing deployment. Update its label/description/importance, replace the champion model, or delete the deployment entirely.
list_model_packages
List model packages in the Model Registry. Model packages are deployment-ready bundles that can be deployed, shared, or used to generate compliance documentation.
create_project
Create a new DataRobot project from a dataset URL or an existing AI Catalog dataset. Optionally sets the target variable and starts Autopilot in one step.
register_model
Register a trained model from a project leaderboard as a model package in the Model Registry. Registered model packages can then be deployed, shared, and used for compliance documentation.
list_projects
List DataRobot projects with optional filtering. Returns project metadata including name, stage, target, and modeling configuration.
manage_project
Update a project's name or delete a project entirely. Use action "update" to rename, or "delete" to soft-delete the project.
get_dataset
Retrieve detailed information about a specific dataset in the AI Catalog including its schema, feature types, size, and processing state.
get_model
Retrieve detailed information about a specific trained model including its type, metrics, training configuration, and optionally its feature impact scores.
Create and manage machine learning projects, train predictive models, and deploy them to production. Run AutoML/Autopilot to automatically select and build best-fit models. Upload and manage datasets in the AI Catalog. Deploy custom models (Python, R, Java) as real-time or batch prediction endpoints. Monitor deployments for data drift, accuracy, and service health. Generate prediction explanations, feature impact scores, and compliance documentation. Manage registered models, credentials, and time series projects. Subscribe to webhook events for project, dataset, and deployment lifecycle changes.
Common questions about connecting Datarobot to AI agents with Metorial.