I'm a double major in Computer Science and Statistics & Data Science at the University of California, Santa Barbara.
I'm mainly interested in networks, systems, and applied research.
A lot of my research starts with an unknown or a large amount of raw system data. I try to turn that into a clear hypothesis, design an experiment to test it, and measure the result using concrete metrics or statistical tests. The goal is to find something useful and reproducible that other people can build on.
I work on network emulation, congestion control, queue management, and application QoE.
Most recently, I worked on a DualPI2 implementation for Mahimahi and compared its behavior against the Linux kernel implementation across different network conditions. This work became a second-author ACM SIGCOMM CCR paper.
I'm also working on experiments that measure how changing network conditions affect real applications like video streaming, conferencing, gaming, and web traffic.
At Lucid Motors, I worked on validation for an in-vehicle AI assistant.
I built agent-based testing systems that generated scenarios, changed conversation state and inputs, and searched for failures that normal deterministic testing could miss.
I'm especially interested in how to measure whether agents are actually useful, whether they behave similarly to humans, and how to detect regressions and recurring failure patterns.
I also work with 3D Gaussian Splatting.
My current work looks at pruning, segmentation, Gaussian lineage, and using agents to help clean and modify reconstructed 3D scenes.
https://illusion-classroom.com
I built Illusion Classroom, an experimental platform for psychology and education research.
It supports large synchronized experiments with hundreds of users and includes real-time communication, video delivery, distributed caching, and experiment management.
I also make music.
