Architecture Framework concepts about AI/ML workloads on Google Cloud

In this section, we will assess how the Google Cloud Architecture Framework can be used with regard to AI/ML workloads on Google Cloud. We will use the steps in the model development life cycle to frame our discussion. As a reminder, the steps in the model development life cycle are summarized at a high level in Figure 13.5:

Figure 13.5: The ML model development life cycle

Let’s begin with the data collection and preparation activities in the model development life cycle, which include gathering, ingesting, storing, and processing data.

Spoiler alert!

You will notice that we have already been using many of these practices throughout this book. Here, we are calling them out explicitly so that you can understand how they apply to workloads in general.

Data collection and preparation

Data management is perhaps the most important of all the topics related to ML governance...