Currently we are reading input from a .csv file, input data should be retrieved from a DB, at this point doesn't really matter which type or where the instance run (because we will be the ones deciding this, eventually) but is to approach a prod scenario.