In a 3scale deployment there's one possible scenario where S3 is used: System's FileStorage can be configured to stored in an S3 bucket.
Additionally, for the backup and restore it's been thought that we could provide the ability to store/restore a backup to/from S3. This is independent of how System's FileStorage is configured.

In both cases interaction with S3 is needed if support for any of the cases wants to be provided now or in the future.

The backup/restore code is executed in a Pod, and Pods use container images.
We have searched for some existing base images in registry.redhat.io that include the 'aws' CLI and we have not been successful finding one.

One possible solution to be able to interact with S3 is creating a new container image (or more depending on how we want to organize it) that includes the needed custom code/tools to interact with S3. If we decide to go towards that approach, doing that implies that we would have to create a new image, productize it, create a GitHub repository to contain the code if we decide to use code, keep it maintained wit changes, CVEs, fixes ...