About Me
I'm a third-year EECS student at UC Berkeley working on geospatial machine learning — using satellite data to answer questions about water. My research splits across two hemispheres: Sentinel-2 imagery over the Maasai regions of Northern Tanzania, where I helped deploy 500 rainwater harvesting units serving 30,000 people, and ICESat-2 altimetry over the polar oceans, where I'm building a Kalman-filter pipeline to track calved icebergs.
The Tanzania work produced a NeurIPS 2024 paper (Machine Learning for Social Impact track) and a pending US patent for dwelling detection in satellite imagery. The iceberg tracker is in active development — you can explore it at /icebergs.
Outside of research, I write poetry and essays at /blog. I contribute to Cal STAR's avionics subteam on the Diablo static-fire engine. I build small hardware projects at /bench.
Research
- NeurIPS 2024 — Machine Learning for Social Impact, High School Track winner
- US Patent Pending 63/703,232 — Dwelling Detection in Satellite Image Data Using a Model
- ML4EO 2024 — Presented at University of Exeter, UK
- NHSJS — Two papers published (Oct 2024, Dec 2024)
Humanitarian Work
Over the past five years I've raised $300K and deployed 500 rainwater harvesting units in the Maasai community of Northern Tanzania, reducing daily water collection time from 9 hours to under 2 hours for 30,000 people. The NeurIPS satellite-ML work grew directly out of this — I needed to know where to put the next 100 units.
Interests
- Writing poetry and short fiction
- Competitive programming and algorithmic challenges
- Hardware prototyping — PCB design, embedded systems
- Improv comedy