A company wants a central catalog for all ML models that currently reside in multiple AWS accounts and are stored in Amazon ECR repositories. What approach will create a central model catalog across accounts?
Choose an answer
Tap an option to check your answer.
Correct answer: Use the Amazon SageMaker Model Registry: create model groups for the ECR-hosted models, create a new AWS account to act as the central catalog using the Model Registry, and attach cross-account resource policies from the original accounts to expose each model group..
Why this is the answer
The Amazon SageMaker Model Registry is designed for cataloging, versioning, and managing ML models, making it the ideal solution for a central model catalog across multiple AWS accounts. By creating model groups for ECR-hosted models and using cross-account resource policies, you can centralize visibility and management. ECR cross-account replication is for replicating container images, not for creating a central model catalog with metadata and versioning. While you could replicate images, it doesn't offer the model management capabilities of SageMaker Model Registry. An AWS Glue Data Catalog is for data, not ML models or container images, and would not provide the necessary model management features.
Pass your exam — without the endless answer hunt
Get every verified question and explanation for this exam in one place, and save hours of prep. 1,000+ certifications · 20+ languages · free to start.
Pass your exam faster → No card needed