Software Engineer | Game Developer | Student
RAYMONDXU
PORTFOLIO
I'm a a Computer Science & Computer Engineering student from Boston. I love building cool projects and learning new things!
My early background focused on programming and developing games in game engines like Unity and Godot. Nowadays, I work in fullstack engineering with interests in AI/ML and embedded systems. I also enjoy creating 3D models and animations in Blender, as well as design websites and graphics with Three.js. I also love competing in hackathons and game jams!
When I'm not on the computer (which is rare), I'm usually running, playing soccer, or rock climbing with friends!
3D-printed brace for stroke and arthritis users with with voice control and grip presets.
Social finance app that gamifies saving; top 5 at HackIllinois 2025.
Unity ML-Agents simulations for space vehicle tasks; won "Interstellar Intelligence" track at BostonHacks 2024.
App visualizing Spotify listening trends with D3.js.
Submission for GMTK Game Jam 2023. 3D Sokoban puzzles but the players and boxes switch roles.
First place/grand prize winner of AngelHacks 3.0. A 2D anime role-playing game about buying assets.
• Optimized agentic video editing pipeline by 40% by redesigning a React Flow-based node editor with pausing, selective restarts, and internal versioning for real-time debugging.
• Built a Remotion-based video editor in TypeScript/Next.js with synchronized timeline and inspector, boosting render performance by 80% and enabling multimodal video agent iteration.
• Built FastAPI endpoints to manage Supabase data and orchestrate Gemini API calls for agent reasoning.
• Selected from 3000+ applicants for 12-month program at MIT including ML coursework with Cornell faculty.
• Partnered with MathWorks on Human Motion Recognition project using IMU data; applied sliding window segmentation on HARTH dataset and trained Random Forest model with 94% accuracy in activity classification.
• Developed Next.js/FastAPI app for real-time activity classification using IMU sensor data from MATLAB app.
• Led a 1-hour weekly live section over a 6-week program to cohorts of 10–15 international students.
• Guided students through Stanford's CS106A-based modules covering Python fundamentals: variables, control flow, functions, lists, and dictionaries, contributing to an overall program completion rate of 90%+ in my section.
Feel free to contact me or check out more of my work on my socials!