hi, i'm jay.

GitHub Email LinkedIn

I study computer science at the University of Washington. I'm also a software engineering intern at Amazon on the Alexa+ project, where my project's AI event tracking features were highlighted at launch. I'm returning to Amazon this summer to work on ML inference to power smarter image search in Amazon Photos.

My career goal is to build systems that are fast, consumer-facing, and algorithmic in nature.

sandbox

jvim Memory-safe modal text editor Arrow
Inspired by kilo, I made a C-based text editor that uses minimal memory and is crash-safe. It's got Vim keybindings and find/replace, and you can try it here.
jpaint OpenGL paint pad for Mac Arrow
MacOS doesn't have a default paint app, so I made one in C. It's got multiple colors (hold your applause), fill, and stroke eraser. Try it here.
data structures Trees and Sketches Arrow
I wrote some toy implementations of various tree-based data structures. They are efficient for a range of problems related to storing and searching data, and you can see them here. I also implemented a couple of data sketches for streaming algorithms, which you can see here.
jumpseat Web-based AI text completion Arrow
A minimalist web-based AI text editor. It's got lots of customizability so I can test different models' capabilities with writing and code completion. Here's the code, and I plan to host it once it's polished.
canvas games Some games on the HTML Canvas Arrow
A collection of simple games implemented on the HTML Canvas. You can play them here.

writing

Signed Graphs June 2025 Arrow
Signed graphs are a way to represent relationships between entities, where edges can be positive or negative. I survey some theorems about their structure and properties, which you can read here.
Sublinear Algorithms May 2025 Arrow
Approximation algorithms often have very simple descriptions but admit strong properties. I give some examples of these "obvious algorithms" in the streaming model, which you can read here.
Sampling Spanning Trees April 2025 Arrow
Given a graph G, can we sample a spanning tree from the uniform distribution of all spanning trees on G? The answer ties together random walks, Markov chains, and linear algebra. Read here.
Quantile Approximation December 2024 Arrow
A presentation I gave to my graduate algorithms class about "Optimal Quantile Approximation in Streams" by Karnin, Lang, and Liberty. Read here.
The computer is useless. It can only answer questions. — Pablo Picasso