This chapter covers
- Building a machine learning model to use in production
- Understanding what APIs are and how they’re helpful
- Deploying a machine learning model
This chapter is written to cover the essential concepts of the job of a machine learning engineer—a person who creates machine learning models and deploys them for the business to use. If your work instead involves creating analyses and reports, it’s easy to be scared of this material. Don’t be! The gap between decision scientist and machine learning engineer is smaller than it seems, and this chapter will be a helpful introduction to the concepts.
Sometimes, the point of a data science project isn’t to answer a question with data; it’s to create a tool that uses a machine learning model to do something useful. Although you could do an analysis to understand what items people tend to buy together, it’s a different task to make a program that recommends the best item to a customer on the website. The work of taking a machine learning model and making it so that it can be used by other parts of the business, such as on the website or in the call center, tends to be complex, and involves data scientists, software engineers, and product managers.
In this chapter, we’ll discuss how to think about how to make models that are part of a product and how to get them off your laptop and into a place where they can function.
Two minor notes before we dive into this topic: