Jupyter is a wonderful tool for interactive computing that is widely being used across the UC system. At UCLA, several JupyterHubs are deployed that allow students and researchers to deploy Jupyter Notebook servers on-demand. These JupyterHubs serve the educational needs of various classes, as well as serving the research needs of various students and faculty. The JupyterHubs variously run on local servers, HPC resources, and cloud-based infrastructure, with containerization allowing for the provision of ready-made computational environments to users.
The ecosystem of Jupyter tools is seeing an increasing number of use cases within research and education at the university. For example, research groups that use sophisticated and custom-made simulation software can use Jupyter tools not only for data analysis, but also to share containers in which their simulation codes are pre-compiled and installed, share notebooks in which their simulation and data workflows are pre-configured, share access to repositories of archived and published data, and share computational resources for closer collaboration. These interactive computing tools can act as a locus that serves a variety of computational and data-driven needs, including serving as an interactive portal to archived research data, methods, and results; serving as a research-educational tool that allows novice researchers to more quickly get up-to-speed in computational research; and serving as a collaboration tool in the fertile but continuously evolving arena of research software design and data analysis. We discuss several particularly illustrative cases at UCLA and offer visions for potential innovation.