I have a large dataset with ~40.000.000.000 records, which is partitioned by date.
To ensure that a single user does not query the entire dataset once, I would like t