About
I'm Trevor — a computer engineering graduate specializing in intelligent systems, robotics, and controls. Currently contributing to cutting-edge marine robotics research at Scripps Institution of Oceanography and preparing to join Typeface as a Software Engineering Intern.
Education
University of California, San Diego
M.S. Electrical and Computer Engineering (Intelligent Systems, Robotics and Controls)
Mar. 2025 – Mar. 2026University of California, San Diego
B.S. Computer Engineering
Sept. 2021 – Mar. 2025- Relevant Coursework: Computer Vision, Autonomous Vehicles, Software Engineering, Circuits and Systems, Design & Analysis of Algorithms, Computer Operating Systems, Computer Architecture
Experience
Incoming Software Engineering Intern, Typeface
- Set to join the Arc Ads team to develop agentic and generative AI systems using TypeScript, React, and Python
- Focus on building scalable, real-time inference pipelines
Robotics Researcher – Embedded Systems & Control, Scripps Institution of Oceanography
- Designed closed-loop controllers in ROS (Python/C++) for 100m-class AUVs, improving real-time energy efficiency by 9% via dynamic depth optimization
- Built embedded autonomy systems using multi-modal sensor fusion (IMU, salinity, depth), reducing manual survey efforts by over 150 hours
- Developed fault-tolerant software architecture enabling resilient deployment across diverse oceanic conditions
- Created GUI for embedded magnetometer data processing, increasing real-time detection accuracy 6× during recovery operations in Vietnam
- First-authored peer-reviewed IEEE Oceans 2024 paper on adaptive autonomy in marine robotics
Frontend Project Manager, Association for Computing Machinery @ UCSD
- Led 13-person team to migrate portal to Next.js & TypeScript, achieving a 250% performance gain via domain-driven architecture and dynamic code splitting
- Built admin dashboard for recruiters using promise-based async ZIP generation, enabling bulk resume download and reducing screening time by 60%
Founder and Full Stack Developer, Notes For Frontliners
- Developed React + Firebase web app enabling real-time note delivery to 800+ users across 14 hospitals, supporting healthcare morale efforts
- Automated community message printing pipeline via custom PDF generator, scaling throughput to 1K+ messages and reducing operational costs
Projects
Autonomous Police Car
ROS2, Python, Computer Vision (YOLO, OpenCV)
Led the software development for an autonomous RC police car project, creating a ROS2 package ('robocar_visual_pursuit_pkg') to enable visual pursuit. Utilized YOLOv4-tiny on an OAK-D camera for real-time vehicle detection and tracking. Implemented adaptive PID steering control, dynamic throttle management based on tracking error, and a real-time parameter tuning interface with persistence for robust performance in chasing scenarios.
VibeCheck
Next.js, React, TypeScript, Hume.ai API, react-vis, react-webcam, react-player, MongoDB, Figma
Proof-of-concept engagement analytics platform using Hume.ai's facial expression models via webcam feed to analyze student engagement during online learning. Provides live feedback graphs for both individual students and aggregate data for teachers to improve teaching methods. Built at CalHacks 2023.
Publications
Chasing Currents: Implementing Depth Optimization for AUV Energy Savings
Trevor Kwan; Raymond Young; Andrew Nager; Mark Anderson; Eric Terrill; Sophia Merrifield