Automation and Real Time Scoring

Your scenarios can be triggered in several ways: using time-based triggers, or when data in a dataset is changed, or even when another scenario finishes.

You can receive updates when a scenario finishes successfully, or only when it fails if you prefer. Find out more about all the different messaging channels that you can use, and how to set them up.

When working with production data that gets updated often, you want to make sure that your data is valid before using it in reports or predictions. In Dataiku DSS you can use metrics to gather information about your data.

You can also define checks on the metrics you computed, letting you interrupt a task if a check fails, and guaranteeing that your data is always safe to use!

If you are using a Dataiku DSS automation Node, you can read this page about importing your bundles so that your project automatically rebuilds in your production environment.

See the deployment to production tutorial to create your first bundle and deploy it to an automation node.

A saved model can be deployed into a Dataiku DSS API node to query a prediction on new data.

The API node provides all the necessary features for scoring in production:

• See the deployment to real-time scoring tutorial to create your first API scoring service and deploy to an API node.
• High availability and scalability for scoring new records.
• Model versioning and rollback using model packages.
• The ability to score in realtime, even with models trained using a distributed engine.
• Scoring multiple models for A/B testing.
• And more advanced capacities such as enriching queries in real-time or handling custom Python models and custom R models.