Summary
In this chapter, we introduced the concepts involved in architecting ML systems, mapped stakeholders, identified common issues and best practices, and outlined the initial architecture. We identified critical building blocks of an ML systems architecture on the data layer and modeling and inference layer. The interconnection between the components was stressed and a specification of features was outlined.
We also addressed how MLflow can be leveraged in your ML platform and the shortcomings that can be complemented by other reference tools.
In the next chapters and section of the book, we will focus on applying the concepts learned so far to real-life systems and we will practice by implementing the architecture of the PsyStock ML platform. We will have one chapter dedicated to each component, starting from specification up to the implementation of the component with practical examples.
