By day, I work in data science + machine learning at Mitsubishi Heavy Industries. Outside work, I'm mainly excited about automating tools with vision and sensing capabilities, & building projects that bring local AI to everyone.
Projects (solo + with friends) include work on clinical report extraction, offline survival, and road damage inference + reporting. These open-source & research projects have received support & recognition from NVIDIA, Anthropic, SAP, & others :)
Provinans
2026With Dr. Blake Gilks
Runner-up, Claude for Science Global Challenge
Using AI to turn scanned longitudinal patient records in The Cancer Genome Atlas (TCGA) into structured datasets for endometrial pathologists.
Featured by Claude
Try it: Hugging Face Space
LifeKit
2026Overall Winner, NVIDIA Spark Hacks
An offline survival guide on the DGX Spark, fine-tuned on FPV datasets to guide users through basic survival skills
Cracked City
2026With Nisten Talraj, Ben Zhou, Efrain Deulofeu
Best Use of Nemotron, NVIDIA Spark Hacks
A fine-tuned DINO model for potholes & road damage, with a severity prioritization system for urban operations
Featured by NVIDIA Developer
Try it: Hugging Face Space