Using the Step Functions Data Science SDK for CI/CD

In November 2019, AWS introduced the AWS Step Functions Data Science SDK for Amazon SageMaker. This SDK allows data science and ML practitioners to programmatically construct Step Function workflows to deliver production-grade ML models. The SDK is designed to be used within a Jupyter Notebook to construct a process that delivers a reproducible ML experiment in the form of a Step Functions workflow, as opposed to reproducing the experiment itself.

Basically, what this means is instead of the ML practitioner exploring data, building algorithms, training models, and evaluating the trained model's performance, they instead construct a state machine to accomplish these tasks automatically. On top of this, the resulting state machine is constructed programmatically, instead of manually defining it with the States Language specification. Therefore, to answer the questions raised in the previous section, the ML practitioner can now...