Chapter 11: Continuous Integration, Deployment, and Training for the MLSDLC

If you review some of the Architecture Best Practices for Machine Learning content, namely the Build a Secure Enterprise Machine Learning Platform on AWS whitepaper, and even the SageMaker documentation on MLOps, you will notice that among the various challenges of automating an application, they all call out the need to have a cross-functional team.

So, why is a cross-functional, agile team so important for automated ML on AWS?

AWS provides numerous ML-related technologies that often overlap in terms of their features to provide their customers with choice and flexibility. Furthermore, the industry provides many tried and tested process guidelines, such as CI/CD, to automate this process. However, neither AWS nor the industry can influence the organizational structure or application development culture of a company. Any changes need to happen within, and done by, the organization.

In Chapter 10,...