Increasing shared understanding of our world by expanding access to geospatial data and machine learning models.

Our initiatives

Radiant MLHub

Radiant MLHub is an library dedicated to open Earth observation training data for use with machine learning algorithms.

Access open datasets and machine learning models from NASA, Planet, University of Maryland and others at


The SpatioTemporal Asset Catalogs (STAC) metadata specification is a common language to describe geospatial information to make it more accessible and interoperable.

Learn more about STAC at

Machine Learning Challenges

We help organize challenges to create capacity building opportunities and accelerate development of machine learning models to interpret Earth observation data.

Learn about our current NASA Harvest Field Boundary Detection Challenge.

Our blog

  • The Naive Origins of the Cloud-optimized GeoTIFF
    Reflections on the emergence of the Cloud-optimized GeoTIFF and how cloud-optimized data can help create a larger and more diverse Earth science community.

  • Celebrating 15 Women Shaping the Future of Earth Science
    In honor of International Women’s Day, we celebrate 15 exceptional women who are shaping the future of Earth Science by pushing boundaries, fostering inclusivity, and advancing a shared understanding of the Earth and its complex systems.

  • Introducing Gina Trapani: Our Newest Board Member
    We are excited to have Gina Trapani join our board, bringing a wealth of experience and expertise from her career in the technology industry. In this Q&A profile, our Executive Director, Jed Sundwall sits down with Gina to discuss her career journey, joining our Board, and her perspective on the potential of the web.