At Tigris we like to think that our object storage platform is a globally distributed multi-tier cache, however when you are doing AI training and other read-heavy workloads sometimes you wish you could have a cache tier on your local network. TAG makes that possible by acting as a read cache for your object storage, including handling authentication.
Think about TAG as a local object storage region. You point your S3 client to TAG and it points to your object storage service of choice (EG: Tigris).
If your object storage workload is read-heavy (AI training, image galleries, or hydrating databases from backups in CI) give TAG a try.
Link on GitHub: https://github.com/tigrisdata/tag
Link to our Discord: https://discord.gg/FS7UMexds (see #tag!)