Metrics Stores vs Feature Stores

In this session, we'll go over what Metric Stores and Feature Stores are, their differences and similarities, why data professionals should learn more about them, and some of their use cases in industry. By the end of the session, you will have a better understanding of how both metric and feature stores might be able to help you and your organisation and, also, how to consume and keep track of the most important KPIs and features that power your data-driven products.

Motivation: With the rise of data-driven applications comes the rise of metrics (churn rate, daily active users, left or right swipes, ads revenue, etc.) and KPIs to keep track of, which makes the process of standardising their definitions across an organisation a potentially complicated task (e.g. different definitions of revenue and active users across divisions of your company). In other words, if the scale of data gave rise to data lakes, the scale of metrics to keep track of have given rise to the Metrics Store. On the other hand, if the rise of data and machine learning-backed applications, especially those operating in complex systems with hundreds if not thousands of features, have highlighted the challenge of keeping track of diverse feature engineering efforts, and this problem has given rise to the Feature Store. As both of these stores solve problems of similar nature but with subtle and yet pronounce differences, let's unpack both of these stores together in this session.Content Level - All levelsWho is it aimed at?
Data professionals at all levels.

Your Speaker
Ramon Perez - Senior Product Developer at Decoded and Developer Advocate at Transform
Ramon works as a senior product developer at Decoded in the APAC region, and also collaborates with Transform as a developer advocate. He was previously a data scientist and educator at Coder Academy in Sydney and before that, he taught statistical programming at London Business School and worked as a research associate at INSEAD. He spends most of his time developing educational data science content for work and for fun, teaching, drinking coffee, mountain biking, speaking at conferences, and exploring many of the outdoor wonders Australia has to offer.

[event ID: 15114]