Every asteroid we fail to detect is a question the sky asked and we couldn't answer. For the past six years, I have been building the instruments to answer it — and I would like to build them with you.
I am writing to apply for the Senior Data Scientist position within JPL's Near-Earth Object Studies group. I recently completed my PhD in Astrophysics at Stanford, where my dissertation focused on the application of deep learning architectures to wide-field survey imagery — specifically, the reconstruction of faint, fast-moving objects from the sort of noisy, cadence-limited data that modern sky surveys produce in abundance.
My published work on Orbital-Net, a convolutional-transformer hybrid for trajectory-aware asteroid detection, was recognized last spring in The Astrophysical Journal and adopted as part of the ZTF reprocessing pipeline. On the held-out B612 benchmark, the model achieved 99.7% detection accuracy on sub-kilometer objects, while reducing false-positive streaks by roughly an order of magnitude over the previous state of the art.
What draws me to JPL specifically is the scale — and the stakes — of the Laboratory's planetary defense mission. The forthcoming NEO Surveyor will generate data volumes that demand fundamentally new approaches to candidate triage, orbit determination, and follow-up scheduling. I would welcome the opportunity to contribute the machine learning infrastructure I have developed for ground-based surveys to a space-based mission with genuine consequence.
Beyond the technical work, I have led a cross-institutional collaboration of eleven researchers across three observatories, mentored four graduate students to first authorship, and taught introductory computational astronomy to more than two hundred undergraduates. I believe the best science happens in teams that take both rigor and kindness seriously.
I have attached my CV and a portfolio of representative publications. I would be delighted to discuss how my work might serve the Laboratory's goals, and I am available for conversation at your convenience.