Building a thriving scientific community: Jupyter as a process, not a product

Presentation given at JupyterCon 2025. 

Abstract:
What does a thriving scientific community look like? To truly thrive, scientific communities must have tools that enable them to access and analyze their varied datasets in one place; platforms that facilitate collaboration across disciplines and expertise levels; and the social structures necessary to connect data, tools, and people into a cohesive, interoperable research fabric. Here, we share insights from the CryoCloud JupyterHub community and broader cross-agency collaborations to illustrate the current state of JupyterHub adoption in a highly dynamic scientific domain. Drawing from user surveys, behavioral patterns, and case studies, we show how researchers integrate the cloud into their workflows, and where they are succeeding or struggling. We also distill core design and community principles, and operational challenges in building sustainable research environments. Finally, we identify ongoing gaps and propose a vision for an ideal, boundaryless scientific community, one where scientific tools and Jupyter infrastructure are not isolated products, but flexible, interoperable processes shaped by ongoing collaboration between tool developers, scientists, and cloud engineers.