So, if I remember right, the endgame was to have an automated ML model running as a service on an AWS instance with SageMaker. Part of that involved having a Jupyter setup installed. Don't remember much past that, though.