Chapter 10: An Introduction to the Machine Learning Software Development Life Cycle (MLSDLC)

At this point in the book, we have reviewed multiple Amazon Web Services (AWS) technologies that can be used to automate the machine learning (ML) process, from automating ML experimentation with Amazon SageMaker Autopilot to automating model training and deployments with AWS CodePipeline, AWS Step Functions, and Amazon Managed Workflows for Apache Airflow (MWAA). We've also seen how various processes can be applied to the task of ML automation by reviewing both a source code-centric and a data-centric methodology to further optimize the ML process. Throughout the previous chapters, we've also seen how different teams within the organization can contribute to the overall success of the ML use case.

In this chapter, we're going to apply what we've already learned, and expand on the key factors that influence a successful execution of an automated, end-to-end (E2E) ML strategy...