Selected work

Projects

Hardware, controls, and software work from coursework, research, and a startup — schematics through firmware through field results.

KiCad ESP32-S3 C/C++ AWS

Co-Founder & Hardware Engineer · Sept 2025 – Present

End-to-end occupancy sensing system using an ESP32-S3 and VL53L8CX ToF depth sensor, achieving >95% counting accuracy at 1–5 m range.
Designed a 4-layer PCB (signal/ground/power/signal) with I2C level shifting, low-noise LDOs, ESD protection, and a ~40-part BOM; achieved first-spin bring-up via JLCPCB.
Embedded C/C++ firmware with temporal filtering cut false positives >30%; streamed data via MQTT to an AWS dashboard at <200 ms latency.

ROS C++ / Python Reinforcement Learning

Undergraduate Researcher, Brown UTRA · Sept – Dec 2025

RL-based attitude/position controllers on Crazyflies improved hover stability in fan-induced downwash ~30% vs. baseline PID in simulation-in-the-loop tests.
Built a telemetry pipeline for state estimation and flight logs, cutting controller iteration time from >1 hr to ~15 min per experiment.
Prototyped multi-UAV disturbance tests; characterized nonlinear coupling with Bode/step-response analyses.

React Django ML recommender

Personal project · District CA-25 Congressional App Challenge winner

Full-stack web app matching students to college majors using an embedding-based similarity recommender.
Built end-to-end with a React frontend and Django backend.

SQL Python

Personal project · co-authored, The Korea Daily

Processed 500k+ census records across 5+ US states to model demographic trends.
Built a reproducible analytics pipeline in SQL and Python; co-authored a published article in The Korea Daily.