Summary

We’ve covered a lot in this chapter, but keep in mind that a lot of the concepts present here will be returned to in subsequent chapters for further discussion. It’s almost impossible to overstate the infrastructure AI/ML will need to be successful because so much of the performance is dependent on how we deliver data and how we manage deployments. We covered the basic definitions of ML and DL, as well as the learning types that both can employ. We also covered some of the basics of setting up and maintaining an AI pipeline and included a few examples of how other companies manage this kind of operation.

Building products that leverage AI/ML is an ambitious endeavor, and this first chapter was meant to provide enough of a foundation for the process of setting up an AI program overall, so that we can build on the various aspects of that process in the following chapters without having to introduce too many new concepts so late in the book. If you’re feeling overwhelmed, it only means you’re grasping the scale necessary for building with AI. That’s a great sign! In Chapter 2, we will get into the specifics of using and maintaining the ML models we briefly introduced earlier in this chapter.