About
The story behind the research.
My Journey
My path into health AI started with a simple question: could a smartphone detect tuberculosis from just a cough? That curiosity led me to develop SCOMAE, a deep learning system that achieved state-of-the-art accuracy in TB detection through cough acoustics. The project earned recognition at Regeneron ISEF 2024, including awards from USAID and Chevron, but more importantly, it showed me the potential of accessible diagnostic tools.
That work caught the attention of Google Research, where I now serve as an External Student Researcher. I pitched the idea of extending audio diagnostics to Google's medical AI efforts, and they gave me a $150k GPU grant to develop audio capabilities for Gemini and MedGemma. It's surreal to see my ISEF foundation model becoming a building block for one of the world's largest AI systems.
I'm currently a freshman at Brown University studying Applied Math and Computer Science, maintaining a 4.0 GPA while juggling research and the occasional game of competitive Catan.
Beyond the Lab
When I'm not training models or debugging code, you'll likely find me in competition. I was a 6x varsity athlete at Taft, competing in cross country, track (300m hurdles), and wrestling (165lb weight class). Sports taught me that consistency beats talent, and that the best preparation happens when no one's watching.
I'm also unreasonably competitive at board games. I hold the all-time ELO world record on Colonist.io (the largest online Catan platform with 4.3 million players) and was the youngest player to qualify for US Catan Nationals, where I placed 30th out of 142.
Economics was another competitive outlet—I captained my high school team to 2nd place at the National Economics Challenge, with our quiz bowl round streamed live on NBC.
What Drives Me
Healthcare is fundamentally an access problem. The best diagnostic tools mean nothing if they're locked behind expensive equipment and specialist appointments. I believe AI can democratize diagnostics—turning every smartphone into a screening device, every rural clinic into a specialist office.
That's why I focus on audio and low-resource approaches. A cough recording costs nothing to capture. If we can make that recording clinically useful, we've just given billions of people access to early detection they never had before.
I'm not interested in AI that replaces doctors. I'm interested in AI that reaches patients who'll never see one.
Education
Brown University
B.Sc Applied Mathematics & Computer Science • 4.0 GPA • Class of 2029
The Taft School
High School • Watertown, CT • Class of 2025
Shanghai American School
Middle School • Shanghai, China