2026
01
eBay
ML Engineer Intern
Built an LLM monitoring workflow and cut false positives 46% with STL decomposition and robust z-score detection.
Open to 2027 full-time ML engineering roles
Hey, I'm Shawn. I like making useful things.
I'm a curious builder based in Ann Arbor. I enjoy untangling hard problems, learning quickly, and working with people who care about the details.
Based inAnn Arbor, MI
Work authorizationF-1 STEM · 36 months OPT eligible
1M+flight legs modeled
2.9Msensor rows analyzed
3 / 80solo competition rank
01 / Experience
From research to production.
Work grounded in noisy data, measurable evaluation, and systems people actually use.2025
02
American Airlines
Applied Data Scientist, student team via Michigan
Flight-delay prediction across 1M+ legs. Beat baselines by 8 AUC points under walk-forward validation.
2022 — 2024
03
Industrial AI Lab, SeoulTech
Research Assistant
Rare-fault detection on a 2.9M-row sensor stream. The work became a Scientific Reports paper.
02 / Selected systems
Built across language, markets, and security.
A compact selection of hands-on ML and systems work.Toolbox
What I reach for to get models into working systems.- Python
- C++
- Go
- SQL
- PyTorch
- Transformers
- XGBoost
- scikit-learn
- pandas
- PostgreSQL
- Docker
- AWS