If it took 40 minutes to insert a 60K line CSV with native postgres using bulk inserts there is something seriously wrong with the code.