Getting started with IBM Orchestration Pipelines
Use IBM Orchestration Pipelines to assemble an automated job for an end-to-end scenario that is part of your data and AI lifecycle. Design your scenario path by dragging programmatic nodes onto a canvas. Refine your pipeline by specifying objects and parameters. Monitor your run results down to each node within the tracker panel.
With IBM Orchestration Pipelines, you can customize your flow for a variety of purposes.
- Gather data, programmatic scripts, and other sources across different projects and workspace to refine and select.
- Run multiple experiments using different services and compare results.
- Create an end-to-end flow by building, running, evaluating, and deploying models or run scripts in a cohesive way.
Checking whether the service is installed
An administrator must install IBM Orchestration Pipelines.
To check whether the service is installed:
- From the navigation menu, select .
- Search for IBM Orchestration Pipelines.
If the service is installed and ready to use, the tile in the catalog shows Ready to use.
Accessing the service
Cloud Pak for Data watsonx™ Depending on the other services that are installed, IBM Orchestration Pipelines is available from either of the IBM Cloud Pak® for Data experience or the IBM watsonx experience.
- Create a new project.
- In your new project, select .
- Search for pipelines and click Automate model lifecycles.
- You can either:
- Create a new pipeline. Select New and enter a name.
- Create a pipeline from sample. Select Sample and click Orchestrate an AutoAI experiment.
Learn more
To learn more about IBM Orchestration Pipelines, see the following topics based on the experience that is available in your environment: