"BRIDGE," a national center opening data science and AI to many fields — NSF AI grant $9M (UC Irvine)

The NSF awarded about $8,998,359 to the BRIDGE National Center, led by the University of California, Irvine (NSF Award 2609582; program: Data Cyberinfrastructure; July 2026 – June 2029).

1A national center widening access to data science

Per the abstract, BRIDGE broadens access to modern data science and AI for scientists across many research fields. Researchers in data-intensive domains — such as criminology, public health, and neurobiology — increasingly need to analyze large datasets and use advanced AI/ML, but often require additional programming expertise, computing infrastructure, or technical support.

BRIDGE eliminates these barriers by providing open, supported, shareable infrastructure that lets researchers build, reuse, and collaborate on data-analysis workflows, datasets, and ML models, and by expanding access to cloud computing so a broader range of scientists can participate in data-driven discovery.

2An open-source, web-based workbench

Technically, the center builds on Apache Texera (Incubating), an open-source, web-based workflow system supporting real-time collaboration, workflow sharing, and elastic access to cloud resources. BRIDGE operates and coordinates Texera-based platforms across scientific communities while providing documentation, tutorials, training, office hours, and support.

Technical goals include more intuitive workflow interfaces, domain-specific AI assistants/co-pilots with privacy and safety protections, mechanisms for protecting data and workflows, automated migration of script-based analyses into reusable workflows, and seamless use of cloud resources for large-scale and GPU-intensive computation — making complex data-science and AI pipelines easier to create, run, share, and reproduce.

3Turning a throwaway script into something shareable

Researchers in fields such as criminology, public health and neurobiology often carry their analysis in scripts they wrote themselves. Those run, but they are hard for anyone else to reuse and they do not scale onto large computation. BRIDGE fills that gap.

1. 1Write the analysis as a scriptIt runs locally, but is hard for anyone other than its author to reproduce
2. 2Migrate it into a workflowA mechanism recasts it into a reusable form
3. 3Collaborate on the webBuilt on Apache Texera, it supports real-time collaboration and sharing
4. 4Run it on cloud resourcesElastic access to large-scale and GPU computing raises the scale

The technical goals also name privacy- and safety-aware domain-specific AI assistants and protection mechanisms for data and workflows. Making things shareable and holding the limits of what may be shared are two faces of one design.