Get hands-on experience with 20+ free Google Cloud products and $300 in free credit for new customers.

Real-time feature engineering

I was wondering if there's a way to do real-time feature engineering with Vertex AI featurestore? An example use-case is if you needed to calculate Z-score for a transaction in real time on a huge number of transactions per second, while also keeping this feature up to date for a quick serving. If this possible, how would it be done? 

0 1 427
1 REPLY 1

For starters, you may check this documentation about the overview of Vertex AI Feature Store. You can use Vertex AI Feature Store to create and manage featurestoresentity types, and features. You can also retrieve feature values in real time to perform fast online predictions.

As to how to use Vertex AI Feature Store on your sample use case, you may review this documentation about Data Model and Resources and check if this fits your purpose.