Move your cursor over the city to sample the contour map, nine points at a time.
Manhattan from Top of the Rock, my photo
About
I'm a Cornell M.Eng Computer Science student focused on backend engineering, systems design and applied optimization. I like turning modeling work, from REST API architecture to PostGIS spatial queries and Gurobi decision models, into infrastructure people can actually use.
Recently I built a geospatial analytics platform on NASA-supported satellite data for Cornell's School of Integrative Plant Science, contributed to a Django and Gurobi system that schedules electric bus charging on real transit data, and shipped backend and mobile features at Splitz and SkyIT Services. I like the unglamorous parts of those systems best, like the cache that keeps a rate-limited API inside its quota, or the transaction that stops a half-finished data load from corrupting a solver run.
I'm most interested in backend and API architecture at scale, optimization-driven decision engines, and financial and operational infrastructure.
Looking for
Full-time backend and full-stack roles
Available
December 2026, after I graduate
Where
New York City and other cities, on-site, hybrid or remote
Strongest in
Python, FastAPI, PostgreSQL, PostGIS, REST APIs
Selected work
Shipped and demoed to farmers
OpenET agricultural water-use platform
Satellite data says how much water a field is losing to the air, but farmers can't easily query it. I built a platform that lets them draw their own field on a map and get answers in plain English.
FastAPI and PostgreSQL/PostGIS backend layered on the OpenET satellite evapotranspiration API, with a Leaflet.js map for drawing field boundaries
Nine-point sampling inside each drawn boundary, averaged into field-level water use with a heatmap overlay
PostGIS cache with query logging that cuts repeat calls to the rate-limited API and keeps the team inside its monthly quota
Seasonal anomaly detection with scikit-learn, plus a Claude-powered chat assistant so non-technical users can ask questions in plain English
Crop type, drought conditions and multi-field comparisons, so farmers can spot unusual water use, irrigation problems and crop stress
Co-presented live to New York farmers at Cornell CALS' 2026 Aurora Farm Field Day with Prof. Yun Yang
Role
Full-stack engineer, Yang Lab
Where
Cornell CALS, with NASA-funded research support
When
April to August 2026
Stack
Python, FastAPI, PostgreSQL, PostGIS, Leaflet.js, Chart.js, scikit-learn, Claude API
New York is moving trash off the curb and into sealed Empire Bins. Our team is building a tool that shows policymakers where that rollout could go wrong and why.
Mapping the feedback loops and tradeoffs between waste generation, curb space, collection reliability and neighborhood equity with the URIS systems-mapping method
Exploring NYC open data (DSNY, 311, PLUTO, the LION street network) to find spatial disparities and pick what the model should predict
Designing the geospatial data pipeline and PostGIS data layer that combine NYC street, building-lot, and 311 data behind a FastAPI backend
Planning an interpretable LightGBM model with SHAP explanations, so a policymaker can see which factors drive predicted risk at each location
Role
ML and geospatial data researcher
Where
AI for Cities group (Dynamic Cities Studio), Cornell Engineering, 5-person team
Electric buses have to charge between routes without blowing the electricity bill. An eight-person team built a system that finds the cheapest schedule that still keeps every bus on time.
Django REST endpoint that bulk-loads real GTFS transit data into PostgreSQL for the Gurobi optimizer
Dataset-prefixing scheme to merge several GTFS datasets into one solver run without key collisions, wrapped in atomic transactions so a failed load never leaves partial data
Time-slot sizing and minimum charging-window constraints in the team's linear program
Friends of mine who make music had nowhere good to sell beats or finish half-written tracks together. I designed and built an iOS app for them on my own as my senior project.
Producers list beats for instant purchase or auction, and artists buy beats and unfinished "opens" to complete
Group-buy lets fans pool money to purchase an unreleased track
Direct messages and requests so artists and producers can find collaborators
Tested whether pretraining a Soft Actor-Critic agent with behavioral cloning speeds up a cluttered object-sorting task. I wrote both training loops, ran 400K-step experiments per condition, and traced a gripper failure to the policy's output distribution. PyTorch, RLBench.
A compiler for a small language of my own, built for Compiler and Interpreter Design.
MyCampusGym iOS app, team of 3, LMU
Find a workout partner on campus, check gym occupancy and classes, log lifts in a workout journal, and post to a campus feed. Shipped as a TestFlight beta. Swift, Firebase.
Cornell Rugby GRIT Team web app
A site my rugby team uses for film days, voting and team notes, plus game schedules, our attack language, solo drills and a workout log. Built with Claude.
Fit and improved logistic growth models for China and India's populations and wrote up the comparison.
Underwriting app for creator contracts Splitz internship, 2025
React Native screens that let underwriters upload contracts, review the data extracted from them, and enter deals, with forms prefilled from JSON so they type less.
Cornell Engineering, Ithaca. September 2026 to now
Urban digital twin for New York's Empire Bins waste-containerization program: systems mapping, NYC open-data analysis and interpretable ML for city policymakers.
Full-Stack Engineer, OpenET Water-Use Platform
Yang Lab, Cornell CALS, Ithaca. April to August 2026
FastAPI, PostgreSQL/PostGIS and Leaflet.js platform on satellite evapotranspiration data, with anomaly detection and a Claude-powered chat assistant. Demoed live to New York farmers.
Full-Stack Engineer, Electric Bus Charging Scheduler
Cornell Engineering, Ithaca. January to May 2026
Django REST ingestion of real GTFS transit data into a Gurobi linear program that minimizes fleet charging cost.
Mobile Engineer Intern, Splitz
Los Angeles, remote. September to December 2025
Built the underwriting module's React Native app from scratch for a fintech startup that scores creator contracts: document upload for PDF, DOCX and images, extracted-data review, and a deal review screen that prefills forms from JSON to cut manual entry for underwriters.
Calgary, remote. May to September 2025. Ranked 2nd of 30 interns.
Built a Puppeteer and pdf-lib export pipeline for the Voop proposal platform, shipped CRUD and CSV-export APIs with Node.js, Express and Prisma on Azure SQL, and fixed a version-restore feature the team had flagged as broken.
M.Eng Computer Science, Cornell University
Graduating December 2026
B.S. Computer Science, Loyola Marymount University
2021 to 2025
Minor in Statistics and Data Science. Dean's List three semesters.
Tools I use
Languages
Python, Java, C, TypeScript, JavaScript, SQL, Swift
PostgreSQL, PostGIS, Azure SQL, geospatial pipelines, API caching
Optimization and frontend
Linear programming, Gurobi, React, React Native, SwiftUI, Leaflet.js
Away from the keyboard
I grew up in Taipei, Taiwan, which explains my two lifelong habits: eating well and playing sports. I played varsity basketball in high school, got into bodybuilding in college, and joined Cornell Rugby when I started my master's.
#6NCR Division II 2026 preseason
Cornell Rugby
I joined the Big Red when I started my M.Eng, and the team is ranked sixth in National Collegiate Rugby's 2026 Division II preseason power rankings. Off the pitch I built the site our team uses for film days and workout logs.
From Taipei
Taipei is home. If you ever want a restaurant recommendation there, from night-market stalls to beef noodle soup, ask me. I speak Mandarin and English.
Basketball, lifting, rugby
Varsity basketball in high school, bodybuilding through college, and now rugby at Cornell. Lifting is also why I helped build MyCampusGym, an app for finding a workout partner on campus.
Gaelic games in Bangkok
In 2025 I joined the Taiwan Celts and traveled with them to the Asian Gaelic Games in Bangkok, where clubs from across Asia compete. We won our championship final.
Travel
I've lived and worked in Taipei, Los Angeles, Shanghai and Ithaca, and Puerto Rico is my favorite place I've ever been. Most of the photos above are from trips like that.
Music
EDM, hip-hop and reggaeton, ideally live at a festival with friends. Music is also why I built Archived, a marketplace for independent artists.
On my watchlist
I've rewatched Friends, How I Met Your Mother and The Big Bang Theory more times than I'll admit, and I'll watch almost anything A24 puts out.