This could be used as an ETL process; one task is to fetch the data, another task to clean and do quality control, and then branching tasks to put the data in i

This could be used as an ETL process; one task is to fetch the data, another task to clean and do quality control, and then branching tasks to put the data in its final destination. Really any task that can be broken into concrete and independent steps could be made into a pipeline for scaling and reliability.