AWS adds S3 Annotations, allowing you to query S3 objects from your Athena Data Lake

Do you have data in S3 buckets you want to put to work training AI models or doing large scale business intelligence?

If so, you will want to check out S3 annotations. It basically allows you to store rich metadata directly attached to the S3 object.

This gets rid of the need to design a separate complex database that associates rich information with S3 paths.

It all comes out of the box with the ability to crawl it with AWS Athena, which, while being an extremely powerful Data Lake tool on its own, also allows you to connect the various tools from Quick Suite.

I could see this leading to some pretty powerful outcomes. For example, if you had a classified data set of specialty images, you could easily train a model to classify or generate more images like these.

The only drawback I see is that you have a bit of vendor lock-in as there likely is not a great way to export these annotations quickly.

Do the positives outweigh the negatives? What use cases do you see for S3 annotations?