About
I'm an embedded systems engineer and student. I write flight computer software at Kairos Autonomi, where I work on GPS-guided autonomous drones. Previously I built air traffic control tooling and a leukemia detection model.
I've been building aircraft since 2024 and writing software for about as long. Most of what I enjoy sits at the seam between the two: firmware that has to survive a real airframe, and the ground software that makes sense of what it sends back.
I'm currently a concurrent-enrollment student at the University of Utah, taking university math alongside high school in Salt Lake City.
Experience
I design embedded control logic for automated, GPS-guided drone systems under the UxV/35 standard, working toward reliable autonomous flight behavior. Most of my day is C++ for onboard flight computers: sensor processing, and the waypoint navigation logic that decides where the aircraft actually goes.
The part I'm proudest of is less glamorous. I built and deployed 30+ target drones for counter-UAS training exercises supporting the United States Marine Corps, which meant pre-flight builds, field testing, and troubleshooting on site to keep exercises on schedule. Firmware is easy to love; a flight line that can't slip is what taught me the most.
Projects
Long-range autonomous FPV fixed-wing aircraft
A fixed-wing airframe that flies GPS-guided waypoint missions on ArduPilot, running on an APM 2.6 flight controller. I integrated an FPV camera and video transmission with onboard telemetry so the aircraft can be monitored in real time while it flies itself.
Navda — AI and automation for air traffic control
Air traffic control still runs on paper strips and aging tower software, against a real controller shortage. Navda is a 3D control tool built around that: live aircraft visualization over Cesium's globe engine, a React and Vite frontend with Tailwind, a Flask backend, Auth0 for authentication, and Supabase holding flight data in real time.
Two pieces went beyond the demo. A watchlist lets a controller track specific flights and get alerted on critical status changes. And flight strips update by voice, using Cerebras AI for low-latency inference with Langchain and Google Speech-to-Text, so the paper step disappears.
Detecting acute lymphoblastic leukemia
A model that reads microscope images of blood cells and sorts them into cancerous and non-cancerous. The interesting work was upstream of the model: curating and cleaning 15,000+ images pulled from several different databases into something consistent enough to train on. It reached 86% accuracy on the held-out test set.
PiHeart — wearable heart rate monitor
A wearable that reads heart rate through a pulse sensor and an ADS1115 ADC, warns through an active buzzer when the rate goes somewhere it shouldn't, and shows live BPM on a 7-segment display. A Ublox NEO-6M GPS module with geopy geocoding broadcasts the wearer's location to emergency services, accurate to 95% in testing.
Education
Concurrent enrollment: Calculus II, Calculus III, and Discrete Math.
IB Diploma Programme. GPA 3.99 / 4.00.
Tools
I write mostly Python and C++, plus JavaScript, Java, HTML, CSS, and SQL. On the web side: React, Flask, Tailwind, Vite, Supabase, Firebase, PostgreSQL, Langchain, Cesium, Cerebras AI, Auth0, PyQt5, Folium, and Git. On the flight side: ArduPilot, PX4, APM 2.6, Matek and ZOHD flight controllers, and Raspberry Pi.
Contact
You can read my résumé, see my code, or find me on LinkedIn. I'm open to internships and collaborations in embedded systems, autonomous flight, and full-stack work — reach out if that's interesting.