Nathan Egbuna

negbuna26 [at] andover [dot] edu

Hi, I'm Nathan, a rising senior at Phillips Academy interested in math, deep learning, and AI applications in finance, medicine, and equity bias.

Currently: Researching how to optimize TTC for models in latent space.


EDUCATION
Phillips Academy
High School Diploma
Relevant courses: Discrete Mathematics, iOS App Development, Object-Oriented Programming in Java

RESEARCH
Optimizing LLMs for Handheld Devices
Quantized medical LLMs for low-power devices, preserving response clarity to expand equitable access to AI health tools.
Bias in Diagnostic Machine Learning Models
Independent research project conducted with guidance from a WPI professor. Developed a prototype to detect overdiagnosis and evaluate accuracy disparities across patient subgroups by analyzing model confidence calibration and error asymmetry.

PROJECTS
VoteNM
Web app that streamlines the voting process by auto-filling New Mexico absentee ballot requests through a user-friendly form and PDF generation.
Univestal
iOS app for youth financial literacy simulating stock and crypto trading with live market data. Available on App Store.
Limit Orderbook Simulator
High-performance simulator of exchange order books with price-time priority matching and randomized trade execution.