Which AWS service is designed to scale on demand to petabytes?

The Amazon Elastic File System (Amazon EFS) is designed to provide scalable file storage for use with AWS Cloud services and on-premises resources. It can scale automatically and on demand to store petabytes of data without requiring the user to provision or manage storage capacity upfront. This makes it an ideal solution for applications that need to handle large-scale data storage and access, as it dynamically adjusts to the storage needs of the application, while providing low-latency access to the stored files.

In contrast, while Amazon QuickSight offers scalable analytics and visualization capabilities, it is not primarily focused on scalable storage. AWS Glue Data Catalog serves as a managed data catalog to facilitate data discovery and management, but it does not directly deal with scaling storage capacity. Amazon SageMaker Batch Transform is primarily designed for running batch predictions at scale but is not a data storage solution like Amazon EFS. Therefore, among the given options, Amazon EFS is best suited for scaling on demand to petabytes of data.