In machine learning, features are attributes used to describe each example. In the process known as feature engineering, transformations are applied to raw data to create features that are suitable for ML models to consume. A feature store is a tool for storing commonly used features, making them available for reuse.

A feature store transforms raw data into feature values by executing automated data pipelines, stores and manages the feature values, and retrieves data for training and scoring. The key benefit of a feature store is that it encapsulates the logic of feature transformations to automatically transform new data and serve up examples for training or scoring. If teams are continually repeating effort to code up feature transformations or copying and pasting feature-engineering code from project to project, a feature store could greatly simplify the overall process and management of features.

In this 30-minute recorded webinar, Thorogood Consultant Mani Singh discusses why feature stores have been increasingly gaining popularity in the industry along with the key benefits, challenges, and limitations of implementing feature stores.

What we cover:

• Introduction to feature stores, their features, and the different cloud offerings that are available
• Live demo showing how feature stores can be used in the MLOps lifecycle
• Key considerations for implementing and using a feature store

Is it for you?

• Are you considering implementing a feature store on a cloud platform?
• Are you interested in learning how feature stores can be used as part of the MLOps lifecycle?
• Are you looking to understand the differences in the offerings across the different cloud platforms?