Columbia University · New York
Medical AI Developer
I build machine learning systems that read medical images and clinical text — retinal scans, brain MRI, discharge summaries — and return something a person can actually act on.
Six studies below
Six things I built, roughly newest first. Ongoing research is marked. Code is on GitHub for everything that has any.
Using diffusion models to generate counterfactual scans — the same patient, one variable changed — to map how a single diagnosis presents differently from person to person.
PyTorch · U-Net · Denoising diffusion · Medical imaging
GitHub
Estimates biological age from a single fundus photograph. The retina is the one place you can see a person's blood vessels without cutting anything open.
PyTorch · CNN regression · Fundus imagery
GitHub
Rewrites discharge summaries, labs and radiology reports at a chosen reading level, and checks every number in the output against the source document.
React · LLM API · Prompt evaluation
GitHub
Classifies tumour type from axial brain MRI slices, with the model's attention rendered back over the scan so a radiologist can see what it looked at.
PyTorch · CNN · Grad-CAM · Augmentation
GitHub
Predicts autism screening outcomes from AQ-10 responses and demographics, with feature weights exposed so the result can be read rather than trusted blind.
Python · scikit-learn · pandas
GitHub
Five-card draw video poker: full hand evaluation, betting and credit state, built as a desktop app in Java Swing.
Java · Swing · OOP
GitHub