Software Engineer · Machine Learning

Jaedon Cox

Computer Science student at The University of Texas at Arlington. I've built full stack from — data pipelines, models, APIs, and front ends.

See the work Start a conversation
01 — SELECTED WORK

Experience and Projects

PRJ-001live in production

Zentrikz

The problem

Sports betting markets are often imperfect, with lines that do not always reflect true player performance or the underlying data. That gap creates opportunities where the probability implied by the market differs from what actually happens.

What I built

An end-to-end machine learning system that automates data ingestion and feature engineering, runs predictive models across multiple player prop types, and generates daily outputs at scale.

PythonXGBoostPandasFastAPISupabaseNext.jsVercelRender
Visit Zentrikz
156,838
batter-games of Statcast data — 7,441 games, 926 hitters
6
XGBoost models — 176 features, SHAP-pruned to 26–52 each
2
scheduled jobs, 0 manual steps — 365 days a year
11,398
lines shipped — 8,171 Python, 3,113 TypeScript, 114 CI
EXP-0012nd overall

Google Hardware Hackathon

Built and troubleshot a Dell server in a hands-on hardware event led by Google team leads. My team finished 2nd overall and received the “Googliest” award, given for the best teamwork of the event.

Dell ServerHardware AssemblyTroubleshootingTeamwork
EXP-00220 tickers

HackUTA

Built a stock market predictor with my team: a neural network trained in TensorFlow to forecast prices across up to 20 different tickers.

PythonTensorFlowNeural Network
Visit Repo
02 — ABOUT

Who I Am

I am a Computer Science student at UT Arlington who likes building things that run on their own. Most of what I have built spans the whole stack — pipelines that pull and shape data, models that make a call on it, an API to serve the result, and a front end to show it. The modeling is the interesting part; the engineering around it is what makes it worth anything.

Machine learning is the field I am aiming at, and I am getting there by building complete systems rather than isolated notebooks. Honest evaluation against real outcomes, not a friendly holdout split, is the bar I hold myself to.

When I am not at my keyboard I am usually watching the Texas Rangers or at the gym. I like staying engaged with sports both as a fan and through an analytical lens, often thinking about how data and performance connect.

Jaedon Cox
based inArlington, TX
studyingCS @ UT Arlington
focusfull-stack systems · ML
stackPython · SQL · TypeScript
statusopen to internships

Open to new problems

Got something that needs building?

Whether it is a pipeline that keeps breaking, a model that looks great offline and falls apart on real data, or a product that needs to exist at all — I would like to hear about it.

jaedoncox7@gmail.com Download résumé