About
Around 2012, I stumbled upon Codecademy and discovered the magic of programming. Starting with basic scripting and game development, I explored web and cloud technologies before discovering my passion for software grounded in hardware experiences. Since then, I've engineered software for tube-launched fixed-wing UAVs, world-record-breaking quadcopters, autonomous racecars, and apple-picking robots.
As I complete my degree at Oregon State, I'm leading SLAM development for Global Formula Racing's autonomous system, researching point cloud-based reinforcement learning in DRAIL, and preparing to rejoin Anduril full-time in Atlanta after graduation. I champion a pragmatic, tool-agnostic approach and am committed to lifelong learning as I explore new technology and industries.
In my free time, I leverage public data to create fun visualizations, like color-matching Pokémon to sports teams. Offline, you can find me cooking international food, building a LEGO, watching a movie, tending to my Animal Crossing island, and hanging out with my cats.
Experience
- Jun — Sep 2025
I architected safety-critical behavior trees for UAV autonomy, contributing foundational mission and fail-safe behaviors. I built embedded Rust middleware for seamless autonomy-autopilot communication and developed a state management system that met strict real-time constraints for coordinate transformations and synchronized data fetching.
- Python
- Rust
- Behavior Trees
- Embedded Software
- Sep 2024 — Present
I work with faculty and grad students on an autonomous agricultural robot that navigates orchards, picks apples, and prunes branches. I've tuned Kalman filters to boost apple localization accuracy and improved reinforcement learning sim-to-real by improving synthetic orchard data generation using Blender. Separately, I've implemented neural networks for point clouds as a preprocessing step for RL training and contrasted their effects on sim-to-real transfer performance for robotic pick-and-place tasks.
- ROS
- Blender
- Gymnasium
- DNNs
- Franka Panda
- Jun — Sep 2024
At this UAV startup, I contributed to SiFly's world-record 3+ hour electric drone flight by delivering software across the entire stack. I built a cloud-based fleet management platform, enabled ML analytics on telemetry data, improved network performance for mission-critical data flows, enhanced WebRTC security, and rapidly developed RESTful APIs. Through proactive teamwork, I helped meet aggressive Alpha launch deadlines.
- Embedded Software
- Django
- WebRTC
- PostgreSQL
- MQTT
- React
- Feb 2024 — Present
As part of an international team, I manage autonomous racecar software projects from concept to integration. I led SLAM and Sensor Fusion integration following a major rework, streamlined observability of system failures, and coordinated with alumni and faculty advisors to prepare our vehicle for global competitions.
- Embedded Software
- ROS
- C++
- Robotics
- Computer Vision
- 2022 — 2023Application Developer, Intern
I developed a web application and RESTful API for monitoring cloud infrastructure in an agile team environment. Through client interviews and customer journey mapping, I drove a significant increase in customer enrollment. I consistently exceeded sprint goals, quickly resolved urgent issues, and effectively communicated progress to stakeholders.
- React
- Node.js
- TypeScript
- AWS
- NoSQL
- Mar — May 2019
As the team lead, I guided a group of interns through the full lifecycle of creating the company website. By setting project timelines, providing technical training, and leveraging analytics tools, I ensured we delivered an engaging site that met business objectives and deadlines.
- HTML
- CSS
- PHP
- Bootstrap
- JQuery
- May — Aug 2018
I taught swimming to children aged 6-16, including those with special needs. I brought energy, patience, and empathy to every lesson, qualities I strive to embody today.
Projects
Autonomous underwater vehicle (AUV) docking policies that maintain performance despite unseen, unpredictable dynamics.
- Isaac Sim
- Deep RL
- Reward Shaping
Combining classical robotics approaches with deep learning to navigate to the correct grid corner by exploring and identifying a hidden handwritten digit.
- Motion Planning
- Deep Learning
- Computer Vision
Autonomous mapping system with a two-tier controller architecture for obstacle avoidance and exploration strategy. Independently developed an expanding wavefront frontier detection algorithm, later discovering it aligned with published research by Phillip Quin et al.
- Motion Planning
- Real-Time Systems
- SLAM
- ROS