Feature management with SageMaker Feature Store
In this section, we will look into what action we might have to take if we were to use a managed feature store instead of Feast in Chapter 4, Adding Feature Store to ML Models.
Important Note
All managed feature stores have a similar workflow; some may be API-based and some work through a CLI. But irrespective of this, the amount of work involved in using the feature store would be similar to what we will discuss in this section. The only reason I am going through SageMaker is familiarity and ease of access to it, using the free trial as a featured product in AWS.
Resources to use SageMaker
In Chapter 4, Adding Feature Store to ML Models, before we started using the feature store, we created a bunch of resources on AWS, such as an S3 bucket, a Redshift cluster, an IAM role, and a Glue catalog table. Conversely, for a managed feature store such as SageMaker, all you need to have is an IAM role that has full access to SageMaker...