Serverless Workflows - Amazon SageMaker Unified Studio

Serverless Workflows

Amazon SageMaker Unified Studio supports serverless workflows powered by Amazon MWAA Serverless. With serverless workflows, you can orchestrate data processing tasks without provisioning or managing Apache Airflow infrastructure. Resources are automatically provisioned when a workflow runs and released when it completes, so you only pay for actual workflow run time.

Serverless workflows in Amazon SageMaker Unified Studio provide the following capabilities:

  • Automatic scaling of compute resources to meet workflow demands without manual intervention.

  • Workflow isolation, where each workflow runs with its own execution role and worker, providing enhanced security and preventing cross-workflow interference.

  • Cost optimization through a pay-per-use model with no upfront costs or minimum provisioning required.

  • Built-in integration with AWS services including Amazon S3, Amazon Redshift, Amazon EMR, AWS Glue, and Amazon SageMaker AI.

  • YAML-based workflow definition files.