Since savepoints are a full snapshot and typically way more expensive than checkpoints, what would be the use cases for automatic savepoints? (as i see on the operator you can take automatic savepoints) I would assume that checkpoints are enough and savepoints should be used in case you want to gracefully shutdown a job and perform some kind of maintainance/upgrade

Especially in enterprise companies you can have requirements like having full backups once every day or more. Don't forget that a checkpoint's lifecycle is managed by Flink, which is not the case for savepoints. I think https://nightlies.apache.org/flink/flink-docs-master/docs/ops/state/checkpoints_vs_savepoints/ has a good overview on this too

one day or more makes sense. I'm was just curious because i believe the default is 5 minutes? for automatic savepoints and i have seen some people making it even less - which seems like a missuse/overkill to me (checkpoints should be enough for such interval) and wanted to get people's thoughts.

I don't think there's any default for checkpointing or savepointing, it's off by default