Chapter 6: Automating the Machine Learning Process Using AWS Step Functions

In the first three chapters of the book, we saw a fundamental process flaw that can impact the automation of an ML use case, namely the handover of a production-grade model, produced by the ML practitioner, to the application development and operations teams. In Chapter 4, Continuous Integration and Continuous Delivery (CI/CD) for Machine Learning, we examined how this issue could be addressed by combining the ML processing into the DevOps process using the CI/CD process methodology.

While that solution inherently addresses the issue, you can also conclude that in terms of overall ownership, the development or platform teams were primarily responsible for building the majority of the final solution. For instance, you will recall from the CI/CD example we used in the previous chapter, the application development teams built the pipeline foundation, as well as the integrations for offloading the data processing...