Skip to content

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

MLOps public preview features

This section provides preliminary documentation for features currently in the public preview pipeline. If not enabled for your organization, the feature is not visible.

Although these features have been tested within the engineering and quality environments, they should not be used in production at this time. Note that public preview functionality is subject to change and that any Support SLA agreements are not applicable.

Availability information

Contact your DataRobot representative or administrator for information on enabling or disabling public preview features.

Available MLOps public preview documentation

Public preview for... Describes...
Service health and accuracy history Service Health and accuracy history allows you to compare the current model and up to five previous models in one place and on the same scale.
Timeliness indicators for predictions and actuals Enable timeliness tracking to retain the last calculated health status and reveal when the status indicators are based on old data.
Real-time deployment notifications Enable real-time notifications for deployments to send alerts about health status changes as they occur.
Model logs for model packages View model logs for model packages from the Model Registry to see successful operations (INFO status) and errors (ERROR status).
Model package artifact creation workflow The improved model package artifact creation workflow provides a clearer and more consistent path to model deployment, with visible connections between a model and its associated model packages.
Automated deployment and replacement of Scoring Code in Snowflake Create a DataRobot-managed Snowflake prediction environment to deploy and replace DataRobot Scoring Code in Snowflake.
Automated deployment and replacement of Scoring Code in AzureML Create a DataRobot-managed AzureML prediction environment to deploy and replace DataRobot Scoring Code in AzureML.
Monitoring jobs for custom metrics Monitoring job definitions allow DataRobot to pull calculated custom metric values from outside of DataRobot into the metric defined on the Custom Metrics tab.
Remote repository file browser for custom models and tasks Browse the folders and files in a remote repository to select the files you want to add to a custom model or task.
Public network access for custom models Access any fully qualified domain name (FQDN) in a public network so that the model can leverage third-party services, or disable public network access to isolate a model from the network and block outgoing traffic.
Runtime parameters for custom models Add runtime parameters to a custom model through the model metadata.
Monitoring support for generative models The text generation target type for DataRobot custom and external models is compatible with generative Large Language Models (LLMs), allowing you to deploy generative models, make predictions, monitor service, usage, and data drift statistics, export data, and create custom metrics.
MLOps reporting for unstructured models Report MLOps statistics from custom inference models created with an unstructured regression, binary, or multiclass target type.
Versioning support in the Model Registry Create registered models to provide an additional layer of organization to your model packages.
Compliance documentation with key values Build custom compliance documentation templates with references to key values, adding the associated data to the template and limiting the manual editing needed to complete the compliance documentation.
MLflow integration for DataRobot Export a model from MLflow and import it into the DataRobot Model Registry, creating key values from the training parameters, metrics, tags, and artifacts in the MLflow model.
Tableau Analytics Extension for deployments Use the Tableau analytics extension to integrate DataRobot predictions into your Tableau project.
Multipart upload for the batch prediction API Upload scoring data through multiple files to improve file intake for large datasets.
Public preview for… Describes...
Service health and accuracy history Service Health and Accuracy history allow you to compare the current model and up to five previous models in one place, on the same scale.
Timeliness indicators for predictions and actuals Enable timeliness tracking to retain the last calculated health status and reveal when the status indicators are based on old data.
Real-time deployment notifications Enable real-time notifications for deployments to send alerts about health status changes as they occur.
Model logs for model packages View model logs for model packages from the Model Registry to see successful operations (INFO status) and errors (ERROR status).
Model package artifact creation workflow The improved model package artifact creation workflow provides a clearer and more consistent path to model deployment, with visible connections between a model and its associated model packages.
Automated deployment and replacement of Scoring Code in Snowflake Create a DataRobot-managed Snowflake prediction environment to deploy and replace DataRobot Scoring Code in Snowflake.
Automated deployment and replacement of Scoring Code in AzureML Create a DataRobot-managed AzureML prediction environment to deploy and replace DataRobot Scoring Code in AzureML.
Monitoring jobs for custom metrics Monitoring job definitions allow DataRobot to pull calculated custom metric values from outside of DataRobot into the metric defined on the Custom Metrics tab.
Remote repository file browser for custom models and tasks Browse the folders and files in a remote repository to select the files you want to add to a custom model or task.
Public network access for custom models Access any fully qualified domain name (FQDN) in a public network so that the model can leverage third-party services, or disable public network access to isolate a model from the network and block outgoing traffic.
Runtime parameters for custom models Add runtime parameters to a custom model through the model metadata.
Custom model proxy for external models (Self-Managed AI Platform only) Create custom model proxies for external models in the Custom Model Workshop.
Monitoring support for generative models The text generation target type for DataRobot custom and external models is compatible with generative Large Language Models (LLMs), allowing you to deploy generative models, make predictions, monitor service, usage, and data drift statistics, export data, and create custom metrics.
MLOps reporting for unstructured models Report MLOps statistics from custom inference models created with an unstructured regression, binary, or multiclass target type.
Versioning support in the Model Registry Create registered models to provide an additional layer of organization to your model packages.
Compliance documentation with key values Build custom compliance documentation templates with references to key values, adding the associated data to the template and limiting the manual editing needed to complete the compliance documentation.
MLflow integration for DataRobot Export a model from MLflow and import it into the DataRobot Model Registry, creating key values from the training parameters, metrics, tags, and artifacts in the MLflow model.
Tableau Analytics Extension for deployments Use the Tableau analytics extension to integrate DataRobot predictions into your Tableau project.
Multipart upload for the batch prediction API Upload scoring data through multiple files to improve file intake for large datasets.

Updated September 18, 2023
Back to top