1 Getting Started with MLOps and ML Engineering
This chapter covers
- Machine Learning life cycle
- Why MLOps is hard
- MLOps + Machine Learning Engineering == MLE
- Building a Machine Learning Platform
- Building Machine Learning Systems
Welcome fellow or aspiring Machine Learning Engineer (MLE)! Machine Learning Operations (MLOps) is still very much a nascent field. Executing successful Machine Learning projects is hard, and you might have heard that most Machine Learning projects end up as failures. Part of the reason for this is due to the sheer complexity, as you shall soon see, due to the many moving pieces.
For the purposes of the book, when we are talking about a Machine Learning Engineer, we're referring to a person that has both ML operations and engineering experience, and this book is going to cover both. This means you'll learn how to:
- Build an ML Platform
- Build and Deploy ML Pipelines
- Extend the ML Platform using various tools depending on use cases
- Implement different kinds of ML services using the ML life cycle as a mental model
- Deploy ML services that are reliable and scalable
These are what we do day-to-day, therefore instead of having a clean separation of both topics, we'd address both, but not call out each of them explicitly as MLOps or ML Engineering specifically.