I build software at the intersection of computer science, quantitative research, and health science.
My background spans biology and computer science, with ongoing graduate work in physiology and computer science. I am especially interested in systems that turn messy information into something testable: data pipelines, research tooling, backend services, quantitative experiments, and focused learning software.
I am also the founder of Quantum Premier Group. Company and research infrastructure is kept separate from the public portfolio where appropriate.
Full-stack Node.js/Express/MySQL operations application for customers, connected devices, firmware releases, deployments, and service tickets. Includes normalized relational modeling, parameterized SQL, Docker-based local setup, screenshots, unit tests, and MySQL-backed integration testing in GitHub Actions.
Test-driven Python engine for Chinese chess, rebuilt from an earlier coursework implementation with a cleaner model/board/game architecture, legal-move validation, check detection, and terminal-state handling.
C implementation of a minimal Unix-like shell covering process creation, I/O redirection, background execution, built-in commands, and signal handling.
Concurrent C TCP client/server project demonstrating sockets, process-based concurrency, message framing, and reversible text transformation.
Small Flask API rebuilt and tested as a focused Python web-service project, with explicit error handling and repeatable pytest checks.
Market-data experimentation sandbox for interactive charting, synthetic OHLCV generation, and lightweight Flask-based exploration. It is intentionally distinct from private quantitative trading/research infrastructure.
- AminoBar — SwiftUI macOS menu-bar amino-acid study utility. The duplicate app entry point has been removed; a macOS CI build and focused data/search tests pass. Interactive menu-bar behavior still needs manual verification.
- Parallel Computing Lab — independently rewritten OpenMP Monte Carlo and optional MPI signal projection with correctness checks. CPU and MPI fixtures pass in Linux CI; reproducible performance measurements remain outstanding.
- Python Learning Archive — curated early Python coursework and learning history.
- RRAMWEBSCRAPE — historical railroad-volume PDF extraction, cleanup, aggregation, and visualization experiment.
- CS340 / CS374 archives — retained for coursework provenance; stronger concepts have been rebuilt into standalone portfolio projects.
- QA take-home — retained as an unfeatured example of pytest/HTTPX API testing against a provided application.
- Fighter — team Unity capstone preserved as a source/design archive with large third-party asset packs separated from portfolio-facing source.
I am continuing to strengthen the engineering around quantitative research, data analysis, testing, reproducibility, and scientific/medical learning tools.
I use AI-assisted development tools as part of my workflow. Repository READMEs distinguish independent work, coursework, team projects, historical experiments, and upstream material so the provenance of each project stays clear.
