A data engineering project involves processing large batches of data on a daily schedule using ETL. The jobs are resource-intensive and vary in size, requiring a scalable, cost-efficient compute solution that can automatically scale based on the workload.

Which compute approach will satisfy the needs described?