Online models generate predictions from prediction requests and these predictions should be logged and made available to (1) evaluate the online performance of models and (2) feature stores to enable the creation of new features and training data.
In this talk, we will show you how you can use the Hopsworks Feature Store as both an Evaluation Store (investigate the performance of models in near real-time and using analytical queries) and a Feature Logging target for prediction requests and prediction outcomes.
What you will learn:
Hopsworks Feature Store is available both on AWS and Azure as a managed platform.
You can register for free, without having to enter payment details.