Part 2 covers the three main workloads a data platform needs to support: processing data, running analytics, and machine learning (ML).

• Chapter 5 discusses processing raw input data into something that better suits our analytical needs. We’ll cover common schemas and see how an identity keyring helps tie the various identities throughout our system together and how a timeline view brings different events together.

• Chapter 6 is all about analytics. It also covers how data engineering can support data science by setting up an environment in which anyone can prototype and deploy analytics to production, while keeping the production environment in good shape.

• Chapter 7 covers machine learning. We’ll see what we need to do to take an ML model that can run consistently and reliably from a Python script to a production pipeline backed by DevOps. Here we’ll introduce Azure Machine Learning and see how it helps us automate these steps.