Governing the ML lifecycle at scale, Part 1: A framework for architecting ML workloads using Amazon SageMaker