A machine learning (ML) engineer has created a feature repository in Amazon SageMaker Feature Store for the company. The company has AWS accounts for development, integration, and production. The comp

Creating an IAM role in the development account (the owner) that integration and production accounts can assume is a secure and standard mechanism for granting cross-account access to AWS resources, including the SageMaker Feature Store and its associated S3 buckets.

Sharing the SageMaker Feature Group directly using AWS Resource Access Manager (RAM) allows other accounts to discover and utilize the feature repository's metadata and online store, enabling direct reuse without duplicating the underlying feature store infrastructure.

AWS Security Token Service (AWS STS) is used to obtain temporary credentials to *assume* an IAM role, but it is not the mechanism for *creating* the cross-account access policy itself.

S3 replication only duplicates the offline feature data; it does not share the SageMaker Feature Group definitions or provide access to the online store, which are essential for feature reuse.

Creating an AWS PrivateLink endpoint provides private network access to SageMaker services from within a VPC, but it does not enable cross-account sharing of the Feature Store itself.