Greetings! I'm Alex.
Seattle-based researcher and open-source enthusiast! Love to talk about board games, machine learning, violin, and graph theory.
Resume Research CV LinkedIn Github Email
Seattle-based researcher and open-source enthusiast! Love to talk about board games, machine learning, violin, and graph theory.
Resume Research CV LinkedIn Github EmailCheck out my Linkedin or read my personal statement for details.
AI for sustainability and Embedded ML at the Ubicomp Lab. Read my research summary for the full story.
I love building solutions to problems my communities face. Most recently, I founded Koel Labs to help foreign speakers master pronunciation.
I play violin with UWSO, build RC planes with Design Build Fly, play tennis and chess, mentor high schoolers, and code up fun projects.
Click the image for a demo of the specific project if available. Click the title for the source code.
Nearly half of all foreign speakers struggle with their accent. Yet previous digital language learning apps focus on vocabulary/grammar and fail to provide the specific and actionable feedback that non-native speakers (like me) need to improve their pronunciation. As CEO and Founder, I assembled a team of 2 other co-founders to rectify this issue. Through the Mozilla Builders Accelerator, we received $75,000 in funding and cloud credits which has enabled us to train SoTA audio models and launch our MVP. We are working closely with top researchers across the US and Canada to iterate on and publish our results.
A commercial version of the attendance scanner made for SYSO coded in collaboration with Claire Li. Quick attendance system powered by QR codes. Manages identity using OAuth2 and Firebase, stores data using SQLite, and has electronic payments/subscriptions using the Braintree payment gateway.
Co-founded a tutoring company with Claire Li. I've taught 20+ students in K-12 topics and music/violin. I also fascillitate peer-reviewed research opportunities for high schoolers. This includes applied projects in the HCI space and projects that involve training custom multi-modal machine learning models. The company website is coded using vanilla HTML/CSS/JS and Firebase for the backend and static file hosting.
Worked with my fellow math major, Austin Ulrigg, to design a novel graph embedding algorithm optimized for cage graphs. This graph family was particularly difficult for previous algorithms because cages have relatively high genus compared to the degree. You can try out a hosted version here. This problem lies at the intersection of graph theory, combinatorics, and topology and has applications in network design, chemistry, infrastructure planning, and more. We have co-authored a paper submitted to Discrete Mathematics.
The eKichabi project is a digital intervention in rural Tanzania to study the economic effects of various mobile technologies to connect farmers and the technical considerations when doing so. I worked with Doctoral Student, Ananditha Raghunath, Dr. Anderson (UW), Dr. Dillon (Cornell), and Dr. Jess Rudder (UChicago) and co-first authored the paper accepted to ACM SIGCHI 2024.
We (Koel Labs) presented our progress developing an MVP for pronunciation learning at Mozilla Builders '24 Demo Day following 12 weeks of intense user validation, machine learning, and HCI research.
Mini-lecture about practical applications of computational linguistics for pronunciation learning for the PhD students at the Tree House seminar.
Introduction to my company, Koel Labs, and our plans for pronunciation learning. Presented to the other 13 cohort teams at the Mozilla Builders Accelerator Kickoff Event in New York.
10 minute in person talk at Para.chi.dub about the eKichabi v2 paper that I co-first-authored.
I was invited to do this 50 minute talk at the CHANGE Seminar for ~40 people based on my research in the ICTD space while working as part of the eKichabi project.
Data processing tools for the University of Washington's Design Build Fly team. Helped our team win 3rd place in the 28th national DBF competition.
A customized assessment software to enhance Cascade Enrichment's K-12 tutoring curriculum developed in collaboration with Claire Li. The tests are commercially available online and rely on UI/UX design, accessibility principles, Javascript, and a Google Form based backend for collecting and autograding responses.
My old youth orchestra (SYSO) needed a way to automate taking attendance for their hundreds of students. I wrote a QR-code based system (hosted on Glitch) that allows people to scan their QR code at a scanning station (any authorized smartphone).
One of the challenges we faced in Science Olympiad was standardizing the way teams test the WiFi antennas they build. So I created a program with a nice GUI to make this easy. It can be downloaded on the Microsoft Store where it now has more than three thousand aquisitions.
This chrome extension enhances toebes.com to allow autogeneration of Cryptographical Puzzles (Code Busters tests for Science Olympiad to help my teammates practice) by simply selecting a type of cipher (affine, ceasar, RSA, Baconian, etc.).
NPM package to automatically manage SQLite database schema changes through a Django-like migration system but framework agnostic and with JavaScript+CLI integration. The library solves the issue of updating a production database schema without losing data or having to manually write/execute SQL queries. Used in production by attendqr.com.
A mini client-side web framework that allows creating webcomponents and managing reactive state with Signals directly in HTML.
Made as an intern at Gooey.AI, this quickly became their top 3 tool with 30K+ uses within the first two weeks and has landed branding deals and social media exposure for the company. It is made using Stable Diffusion models with Controlnets (deployed using Docker and Kubernetes) and also involved making a URL shortener using Django.
I wanted to see if I could develop and deploy a small web application in a few hours (spread across 5 days) without any frameworks, nor referencing my old projects, and keeping google searches to a minimum. So I challenged myself to do it with this simple Tierlist Maker. The only thing I had to google was the compression library and icon API as I had never used those before (and these were non essential features anyways). As an added bonus, it is completely ad free and runs locally in the browser.
A classic web-based Tetris game with no ads. Has PWA capabilities so it can be downloaded and used offline. Simple learning project to explore game-loops, touch controls, and online leaderboards using a hosted database.
A word game inspired by the popular Wordle but with ocean themed background graphics. Coded in collaboration with Claire Li. A live demo of a similar game I coded on my own can be found here: lawdle.web.app (the challenge is to guess the state/country of the absurd law, or in a different mode select the AI generated law)
The challenge here was to write a 3D rendering engine completely from scratch (down to the drawing of individual pixels) using only the CPU (so no help from the graphics card). This project taught me a lot about the math behind projection matrices, lighting, 3D transformations, and optimizing algorithms (parallel triangle drawing routines, frustrum culling, and more).
This Science Olympiad challenge involved building and programming a sensor and its circuitry from scratch (using only basic circuit components and DIP) to measure salt concentration in water samples. I used a Raspberry Pi microcontroller to communicate via the I2C protocol to a NAU7802 ADC chip reading a voltage divider circuit. With proper callibration, probe submersion, and luck it is capable of less than 1% measurement error.
Reach me via alex@sandergi.com or scan the QR code to add me to your contacts.
