Possibly the original data is logically compressed. E.g. payer A pays 110% of our standard rates, payer B pays 85% of our standard rates. Those two rows could t

Possibly the original data is logically compressed. E.g. payer A pays 110% of our standard rates, payer B pays 85% of our standard rates. Those two rows could translate into thousands of CSV lines depending on the number of procedures. Maybe you have a couple one off negotiations for high volume procedures, but even still the source data could be several orders of magnitude smaller than the dumps.