Workflow management, serverless, and cloud: How to do it better

Getting more out of workflow management in a serverless environment

Workflow management is often a thorny problem, and it doesn’t get easier with a move to the cloud. In fact, different cloud-provider requirements and languages may exacerbate workflow management issues. In this episode of the podcast, Mike Kavis and guest Cohesion.Dev’s Soam Vasani discuss workflow management in a serverless environment. They take a look at how companies are using workflow management to enhance AI/ML and how to execute the process better, no matter where it occurs. Mike and Soam dissect the advantages and issues with different tools and techniques, as well as cloud-specific workflow management topics, and Soam gives his take on how companies can make the process easier, faster, and more effective.

Disclaimer: As referenced in this podcast, “Amazon” refers to AWS (Amazon Web Services) and “Google” refers to GCP (Google Cloud Platform).

If you think about a workflow and the idea of saying, "Do this, do that, if this happens do something else," that's programming, right? The idea of expressing that series of tests, that is programming. And any complex enough workflow starts looking like the flowchart of a program.

Soam Vasani leads Cohesion.Dev, a workflow compiler for AWS. Previously, he led the development of the Fission FaaS for Kubernetes, and the Fission Workflow system for Kubernetes.