I have a process in Airflow that reads some CSV to a Pandas DF and then it does some operations with it, and save the results as parquet files in another bucket in S3.
Th