Building a campus-wide platform for Data Science / Machine Learning coursework

Over the past two years, with input and vision from researchers building high-performance systems such as the Pacific Research Platform (PRP), we have built out a hybrid cloud/on-premises environment for student Data Science and Machine Learning, supporting both coursework and student-focused independent study research.  The system is now used by 1300 students per quarter across multiple departments and disciplines, from Engineering, to Cognitive Science, to Visual Arts.   We will share technical and non-technical lessons learned in the deployment and operation of customized per-course configurations of JupyterHub, Kubernetes, and on-prem GPU hardware.


Session skill level
Session Track
Enabling Teaching and Learning