Introducing MLOps

Chapter 3. Key MLOps Features

Mark Treveil

MLOps affects many different roles across the organization and, in turn, many parts of the machine learning life cycle. This chapter introduces the five key components of MLOps (development, deployment, monitoring, iteration, and governance) at a high level as a foundation for Chapters 4 through 8, which delve into the more technical details and requirements of these components.

A Primer on Machine Learning

To understand the key features of MLOps, it’s essential first to understand how machine learning works and be intimately familiar with its specificities. Though often overlooked in its role as a part of MLOps, ultimately algorithm selection (or how machine learning models are built) can have a direct impact on MLOps processes.

At its core, machine learning is the science of computer algorithms that automatically learn and improve from experience rather than being explicitly programmed. The algorithms analyze sample data, known as training data, to build a software model that can make predictions.

For example, an image recognition model might be able to identify the type of electricity meter from a photograph by searching for key patterns in the image that distinguish each type of meter. Another example is an insurance recommender model, which might suggest additional insurance products that a specific existing customer is most likely to buy based on the previous behavior of similar customers.

When faced with unseen data, be it a photo ...