SageMaker Workflows - Amazon SageMaker

SageMaker Workflows

You can manage your Amazon SageMaker training and inference workflows using Amazon SageMaker Studio and the Amazon SageMaker Python SDK. With the available tools, you can simplify your SageMaker process and integrate it into your existing project.

The following workflow technologies are supported.

  • Amazon SageMaker Model Building Pipelines: SageMaker's tool for building and managing end-to-end ML pipelines.

  • Airflow Workflows: SageMaker APIs to export configurations for creating and managing Airflow workflows.

  • Kubernetes Orchestration: SageMaker custom operators for your Kubernetes cluster, as well as custom components for Kubeflow Pipelines.

  • AWS Step Functions: Create multi-step machine learning workflows in Python that orchestrate SageMaker infrastructure without having to provision your resources separately.

For more information on managing SageMaker training and inference, see Amazon SageMaker Python SDK Workflows.