Isaac Corley is Director of AI/ML Research at Taylor Geospatial, where he leads research and engineering on Earth observation foundation models and open data products. He has spent the last decade working across sides of the field that rarely overlap — production ML at Wherobots and Zeitview, applied research at Microsoft AI for Good, and PI work on the IARPA SMART program at BlackSky — and tends to focus on the places where remote sensing benchmarks and operational deployment fail to agree.
He publishes regularly at CVPR, NeurIPS, ICML, ICLR, and SIGSPATIAL, with several best-paper recognitions, and is a core maintainer of TorchGeo and other widely-used open-source libraries that sit underneath much of the field’s tooling. Much of his recent work has focused on evaluation methodology and reproducibility in geospatial foundation models.
