Cornell University
B.A. in Computer Science and Mathematics
Software Developer
Ethan Ngai
Building production web apps
Undergraduate at Cornell University studying Computer Science and Mathematics. I build software across web, backend, XR, and machine learning.
About
Learning by building real things
I love designing software that helps people in their everyday life. That includes full-stack applications for education, mixed reality productivity apps, and machine learning tools that make life easier.
At Cornell, I balance coursework, clubs, and project teams. I am on the software team at Cornell XR and currently a TA for CS 2110. Outside school and work, I , , and play .
Work Experience
Building beyond localhost
Teaching Assistant
Cornell Bowers College of Computing and Information Science
Jan 2026 – Present
Ithaca, NY
- Leads weekly CS 2110 discussion with 20+ students on data structures and algorithms.
- Grades assignments and exams, and runs office hours for concept clarification and debugging.
- Develops Java autograder logic and test cases for course management software.
Software Team Member
Cornell XR Project Team
Nov 2025 – Present
Ithaca, NY
- Developing persistent memory framework for smart glasses with conversation transcription and facial recognition.
- Creating RAG-based semantic recall with pgvector, FastAPI, SQLAlchemy, and OpenAI API.
Machine Learning Research Assistant
VCU College of Engineering Research Lab
Jul 2024 – May 2025
Richmond, VA
- Researched machine-learning methods for wireless communication sensing.
- Built Python workflows that mapped WiFi channel-state information to Quest 3 depth imagery.
- Trained PyTorch denoising models and improved simulated-signal accuracy by nearly 40%.
Skills
Tech in my toolbox
Languages
- Java
- Python
- C#
- OCaml
- C/C++
- JavaScript
- TypeScript
- HTML
- CSS
Frameworks
- Spring Boot
- React
- Node.js
- TanStack Start
- Next.js
- FastAPI
- Flask
- SQLAlchemy
- PyTorch
Tools
- PostgreSQL
- Docker
- Cloudflare
- Supabase
- Convex
- Vercel
- Vite
- Claude Code
- Git
- Unity
- AWS
Projects
Hackathons and late nights
Sharpnd
Adaptive studying and tutoring platform focused on personalized question selection and modern edge deployment.
Built with TanStack Start, React, Drizzle ORM, and PostgreSQL. Created a CI/CD pipeline for Cloudflare Workers, Neon, and R2 storage, designed an adaptive question algorithm using semantic vector similarity with tag-based fuzzy search, and implemented authentication and email flows with Better Auth and Google OAuth.
- TanStack Start
- React
- Drizzle ORM
- PostgreSQL
- Cloudflare Workers
- Neon
- R2
- Better Auth
- Google OAuth
IPArena
Competitive and solo quiz game for IPA transcription practice.
Built with Spring Boot services, JPA-backed PostgreSQL, Dockerized deployment, and a React client.
- Spring Boot
- PostgreSQL
- Docker
- React
- Java
Critter World Simulation
Collaborative language parser and distributed simulation platform from CS 2112.
Implemented custom-language recursive descent parser and AST, simulation logic, multithreaded distributed computing, SparkJava APIs, and JavaFX client GUI.
- Java
- SparkJava
- JavaFX
RSA Encryption Toolkit
Cipher suite with Caesar, substitution, Vigenere, and full RSA support for large-file encryption.
Included strong JUnit coverage and a command-line workflow for secure file operations.
- Java
- JUnit
- Cryptography
SafeStep
Hack@Brown 2026 project for safer, accessible route planning across walk and drive travel modes.
Built a Spring Boot backend with PostgreSQL/PostGIS, integrated A* pathfinding with hazard-aware edge weights, and shipped a React + Leaflet/OpenStreetMap client with crowdsourced image reporting and Gemini-assisted danger scoring. Containerized with Docker Compose.
- React
- Spring Boot
- PostgreSQL
- Docker
- Leaflet
- OpenStreetMap
- Gemini API
Spinder
Big Red Hacks 2025 project that makes music discovery feel like a dating app.
Led backend architecture development with Flask RESTful APIs, Gemini prompt engineering, and SQLAlchemy persistence for personalized playlists.
- SwiftUI
- Flask
- Spotify API
- Gemini
- Python
Home Archery
Quest 3 mixed-reality archery training range with room-aware boundaries, accurate arrow physics, and precise hand tracking.
Designed and developed with Unity, C#, and Meta XR SDK for immersive gameplay right in your bedroom.
- Unity
- Meta XR
- C#