EECS @ UC Berkeley
Undergraduate researcher in machine learning and robotics, currently building autonomous surgical debridement technology at Berkeley AUTOLab.
Berkeley, California
01 — About
I am an undergraduate at UC Berkeley studying Electrical Engineering & Computer Science. Go Bears!
My interests lie in machine learning and robotics. As of Spring 2026, I joined AUTOLab, where I work on medical imaging technology for sutures and benchmark autonomous surgical debridement pipelines.
Welcome to my little corner of the internet! 🫶
Semester by semester, plus the research groups I was part of at the time.
02 — Work
A few things I've built. Each one taught me something I still use.
I work on medical imaging technology for sutures and benchmark autonomous surgical debridement pipelines — measuring how reliably a robot can identify and remove damaged tissue, and where those pipelines still break down.
At UC Berkeley's GamesCrafters Research Group, I contribute to GamesMan, a system for solving and playing two-person abstract strategy games perfectly. I implemented and fully solved game variants — most notably 9 versions of Hexapawn — storing solutions in a high-performance database. The team improves system efficiency with game-specific optimizations, including hash functions, while developing intuitive GUIs and new variants for interactive learning and strategy analysis. The work sits at the intersection of algorithmic game theory, software engineering, and educational tools.
Implemented and fully solved 9 variants of Hexapawn
A handwritten-digit classifier written in RISC-V assembly. Matrix multiplication, ReLU and argmax are implemented from scratch, so the whole neural network runs on hand-managed memory — a lesson in how much the abstractions above actually cost.
A speaker classification system built on SVD and PCA for dimensionality reduction, projecting recorded audio into a lower-dimensional space where the classes separate cleanly.
85%+ classification accuracy
A Java backend for exploring historical word frequency in the Google NGrams dataset. I implemented the TimeSeries and NGramMap classes that make the queries efficient, plus the plotting layer that turns them into readable graphs.
A Scheme interpreter in Python with environment frames, symbol lookup and user-defined procedures, using the trampoline technique for tail-call optimization so deep recursion doesn't blow the stack.
Spring 2025
Monte Carlo simulation estimating percolation thresholds in a grid using weighted quick-union, with a visualizer for debugging.
October 2024
Microphone localization from cross-correlation and least-squares estimation of time-of-arrival differences.
October 2024
A turn-based tower defense game for CS 61A, built around an object-oriented model of ant placement and bee behavior.
September 2024
Spectrogram-based song identification that matches a brief recorded clip against a fingerprint database.
2023
Java robot programming with reinforcement learning to tune combat strategy over repeated matches.
2023
Edge detection and pixel-level image processing algorithms written from scratch.
September 2021
Pathfinding and search algorithms implemented for a UC Berkeley outreach project.
October 2022
A utility that converts an entire book's text to Pig Latin, line by line.
03 — Engineering
Before EECS, I spent high school in robotics and aerospace — competition teams, spacecraft design, propulsion testing, and a lot of things built to see whether they'd hold.
Served as team president in 2023, restructuring the club into separate building and educational sessions. The team earned the Judge's Award at the August 2023 regional competition and qualified for both the state championship and the CREATE US Robotics competition.
Led a team of seven through a full Mars mission design — launch mechanics, planetary research, crew health protocols and a return trajectory built on Hohmann transfers and orbital mechanics.
Measured impulse data from Estes A8-3, B6-4 and C6-5 motors using force sensors and test equipment, then analyzed how impulse affected rocket performance.
Designed a vehicle from spring potential energy and friction calculations, and won Best Mousetrap Car Design of the Year.
Designed and simulated a walking robot with a partner using Arduino, servo motors and CAD, incorporating kinematics and a localization system.
Tested foam and fiberglass combinations on hydraulic equipment, measuring deflection and load-bearing capacity for aerospace applications.
Instrument Landing System interpretation, VOR navigation, aircraft instrumentation, and scenarios run in NASA's Air Traffic Control simulator.
Three design iterations optimizing nose mass and wing dimensions, reaching a maximum distance of roughly 72.5 feet.
Designed airfoils to meet specific flight requirements across altitudes using NASA's airfoil simulator alongside spreadsheet analysis.
At the Explore Engineering Innovation Program, built The Active Susan: a CPX board device using light sensors and an accelerometer to prompt exercise breaks.
An automated sorting system built with a partner from sensors, motors and 3D-printed parts, combining object detection with piezoelectric crystals for material sensing.
An articulated robotic arm built in Honors Principles of Engineering using inverse kinematics, torque calculations and sensor-based control.
A plywood vehicle powered by hydrogen fuel cells and solar panels, travelling up to 100 feet across several wheel and structure configurations.
04 — Principles
The handful of ideas I keep coming back to.
I love my family and am super thankful for the life I am given. My family, both biological and by faith, mean the world to me.
As the Navy says, pain is weakness leaving the body.
A working thing in someone's hands teaches more in a week than a spec teaches in a month. Iterate in public.
One careful decision doesn't matter. Fifty of them are the difference between work that runs and work people can build on.
I am an avid runner, triathlete, and reader.
“Man reading should be man intensely alive. The book should be a ball of light in one's hand.”
I ran my first half marathon in 2019 and have been pushing my capabilities ever since.
I race with UC Berkeley's Competitive Triathlon team.
The team has taken me to Santa Barbara, Davis, Cal Poly, and Florida to compete.
Trails and elevation, whenever I can get to them.
I've always loved literature — Dostoevsky, Kafka and Hemingway most of all.
I am working toward 3/4 of the Modern Library's Top 100. Most recently finished #74, A Farewell to Arms.
05 — Contact
I am always glad to talk about research, machine learning and robotics, or new opportunities. The fastest way to reach me is email.
06 — Beauty
Here are some things I have found beautiful of late.
