How would you design an automated reconciliation process to ensure that data ingested matches data processed at the end of the pipeline?
