Models that run on your own infrastructure (outside of DataRobot) may be run in different environments and can have differing deployment permissions and approval processes. For example, while any user may have permission to deploy a model to a test environment, deployment to production may require a strict approval workflow and only be permitted by those authorized to do so. Prediction environments support this deployment governance by grouping deployment environments and supporting grouped deployment permissions and approval workflows. You can create, manage, and share prediction environments across DataRobot. This allows you to specify the prediction environments used for both DataRobot models running on the Portable Prediction Server and remote models monitored by the monitoring agent.
Add an external prediction environment¶
To deploy models on external infrastructure, you create a custom external prediction environment. Navigate to the Deployments > Prediction Environments page and select Add prediction environment.
Complete the fields:
Name: Enter a descriptive prediction environment name.
Description: (Optional) Enter a description of the external prediction environment.
Platform: Select the external platform on which the model is running and making predictions.
Model type support
Different environments may support different model types: DataRobot-built, Scoring Code, custom, or external. When an administrator creates a prediction environment, they select which model types are supported.
Once completed, click Add environment. The environment is now available from the Prediction Environments page.
Edit an external prediction environment¶
To edit the prediction environment details you set when you created the environment and to assign a Service Account, navigate to the Deployments > Prediction Environments page and click the row containing the prediction environment you want to edit:
Name: Update the external prediction environment name you set when creating the environment.
Description: Update the external prediction environment description or add one if you haven't already.
Platform: Update the external platform you selected when creating the external prediction environment.
Service Account: Select the account that should have access to each deployment within this prediction environment. Only owners of the current prediction environment are available in the list of service accounts.
DataRobot recommends using an administrative service account as the account holder (an account that has access to each deployment that uses the configured prediction environment).
Select a prediction environment¶
After you add a prediction environment to DataRobot, you can deploy a model and use the prediction environment for the deployment. Specify the prediction environment in the Inference section.
Once an environment is specified and the deployment is created, the prediction environment cannot be changed.
The sharing capability allows appropriate user roles to grant permissions for prediction environments.
When you have created a prediction environment and want to share it with others, select Share () from the dashboard.
This takes you to the sharing window, which lists each associated user and their role. To remove a user, click the X button to the right of their role.
To re-assign a user's role, click on the assigned role and assign a new one from the dropdown.
To add a new user, enter their username in the Share with field and choose their role from the dropdown. Then click Share.
This action initiates an email notification.