Feature Store for Machine Learning

Section 2 – A Feature Store in Action

This section concentrates on the what? and how? aspects of feature management in Machine Learning (ML). In this section, we will start with an introduction to an open source feature store, Feast, followed by different terminologies and basic API usage. We will reuse the same ML problem that we discussed in Section 1, Why Do We Need a Feature Store?, create a Feast infrastructure on AWS, and include it in our ML pipeline. This inclusion enables us to look at how a feature store decouples our ML pipeline into different stages and the changes it brings to model training and inference. Once the model development is complete, we will look at how the capabilities of a feature store make it easy to move a model ...