I build machine learning and analytics products that are meant to be used, not just trained and then abandoned in a notebook.

My work usually sits at the intersection of:

• applied ML and forecasting
• evaluation, calibration, and decision support
• FastAPI, Python, SQL, and TypeScript
• CLI tools, dashboards, and product-style workflows

I care a lot about shipping systems that are inspectable, measurable, and useful in the real world.

I like models, but I like working software, clean dashboards, and honest evaluation a little more.

Sports betting platform with ML predictions, Kelly-based bet sizing, live odds views, and product-style delivery. Strong signal for end-to-end application work.

Applied modeling repo for NBA and NFL edge detection using logistic regression, XGBoost, and ensemble methods.

Monorepo that combines ai-advantage and sports-betting-ml as apps/* directories with subtree-imported history for coordinated evolution.

Reusable ratings and win-probability library with Kelly helpers. Good example of turning modeling logic into a usable package.

TypeScript package for Kelly sizing, odds conversion, and bankroll math. Lightweight and package-oriented.

FastAPI and Chart.js dashboard for calibration, rolling accuracy, and CLV reporting. Original repo archived; code preserved on oss-archive.

CLI that scores repository quality across README, CI, licensing, staleness, and maintenance indicators. Original repo archived; code preserved on oss-archive.

Hub repo: one branch per retired public project (archive/<repo-name>) copied before those repos were archived read-only. Branch index.

Languages: Python, R, SQL, TypeScript, JavaScript, Bash, HTML/CSS

ML/Data: scikit-learn, XGBoost, pandas, NumPy, model evaluation, calibration, backtesting, forecasting

Backend and Product: FastAPI, SQLite, REST APIs, CLI design, dashboards, reporting pipelines

Workflow: GitHub Actions, reproducible tooling, documentation, repo health, developer experience

GitHub layout: 9 active public repositories plus oss-archive (frozen copies of 24 retired repos as archive/<name> branches before archiving originals). Public toolkit matches this story.

Retired OSS (odds tools, eval pipeline, coursework, agents, etc.) is read-only archived on GitHub but browsable via oss-archive branches or all repositories including archived.

• B.S. Information Science, University of South Florida, expected May 2026
• Starting M.S. Artificial Intelligence at the University of South Florida in August 2026
• Interested in ML Engineer, Data Scientist, AI Engineer, and applied research roles
• Open to remote-first opportunities

I write about ML systems, analytics, and applied AI on Alloway AI.

If you like the open-source work and want to support it:

• GitHub Sponsors
• Crypto wallet: 0x6F278Ce76BA5ED31Fd9bE646D074863e126836E9