Skip to content

Apply for Funding

Current and past funding opportunities

What we fund


We target software that serves as foundational infrastructure for data-intensive science and AI-driven discovery. Rather than funding AI models themselves, we invest in the open source software stack that models depend on, ensuring it is robust, interoperable, and sustainable. See individual calls for proposals for more on what we fund.

Data management and representation

Tools for representing, managing, curating, and structuring scientific data for use in model training. These are the foundation AI models are built on.

Model training and evaluation

Infrastructure for reproducible training, benchmarking, and validation of scientific AI models.

Hardware acceleration and HPC

Libraries and tools that help scientists take advantage of GPUs, TPUs, and high-performance computing environments.

Agentic workflows and automation

Workflow systems and APIs that support autonomous, agentic scientific experimentation at scale.

Header Video: Cell trajectories in a developing zebrafish embryo tracked by Ultrack — an open source library from the Royer lab at Biohub for large-scale cell tracking under segmentation uncertainty. Data by Jordão Bragantini from an experiment by Xiang Zhao, video by Alexandre Dizieux.