We have a few applications running on AWS EMR and Spark. Each of these applications run either streaming mode or in batch mode. We have CFT for EMR and we are using concourse pi