I just finished a data science internship at World Wide Technology building an end-to-end LiDAR-to-CAD pipeline, and I'm now writing my MS thesis on predicting MCI-to-Alzheimer's conversion using partial AUC maximization as the training objective. I care about applied AI and building things that matter.
Selected Work
Built an end-to-end ML pipeline converting LiDAR point cloud buildings and road edges into engineering-grade CAD models, projected to cut extraction labor by 50%+ and generate $12.5M in annual revenue. Also placed 2nd of 19 teams presenting CARDINAL, an intern operations platform, to 100+ people.
Honors thesis on predicting whether MCI patients will convert to Alzheimer's Disease from 3D brain MRI, with the core question of whether partial AUC can be optimized directly as a loss, ultimately targeting the clinical metric doctors actually use.
Forecasting pipeline built from millions of bag-scan records to predict baggage load curves two weeks out from flight-schedule features, enabling proactive ground staffing for American Airlines.
Simple, healthy habits make a complex life manageable.