Poiesis, TES, and Its Applications

At its core, Poiesis allows you to describe compute tasks over inputs using a standard schema, then schedules and executes those tasks within a containerized environment.

Think of it this way: if your problem can be reduced to "run this command with these inputs and get the outputs," then Poiesis can help you run it—at scale, across systems, and with Kubernetes-level isolation.

Why TES?

TES provides a standard interface to describe computational tasks, making it easier to plug into different infrastructure without having to rewrite execution logic. By implementing TES on Kubernetes, Poiesis brings this standard to the cloud-native world, enabling seamless integration with:

• Bioinformatics pipelines
• Federated and multi-tenant platforms
• Workflow engines like Cromwell, Nextflow, or Toil etc
• Custom UIs or APIs that trigger compute jobs

Poiesis as a Federated Task Layer

You can think of Poiesis as a task execution abstraction that sits between a requestor (user, workflow engine, or system) and the Kubernetes cluster. It decouples who asks for the work from where and how it's run. This is especially powerful for:

• Multi-user scientific computing platforms
• Shared infrastructure in research organizations
• Data access governance (e.g., secure processing of protected datasets)
• Workload bursting across clusters or cloud providers

What’s Ahead

The following pages will walk through real-world use cases where Poiesis can serve as the intermediary layer to simplify compute orchestration in federated or scalable systems.