- Saving frequently used data transformations for reuse
- Tracking the cost of individual Glue ETL jobs
- Marking specific points in an ETL workflow for debugging
- Persisting state from a job run to prevent reprocessing of old data