AI Agents
Learning systems that plan, use tools, and adapt over extended interactions.
Personal research portfolio · Software Engineering
WANG YUZHE
Undergraduate Exploring AI Agents, World Models, and Embodied Intelligence
I am a Software Engineering undergraduate at Xiamen University Malaysia, interested in building and studying intelligent systems that can reason, model environments, and act in the physical world. My current focus spans AI agents, world models, and embodied intelligence, grounded in algorithms, systems, and embedded implementation.
Open to research opportunities
01 / Current direction
Building toward open-ended intelligent systems
I am seeking research opportunities where I can contribute strong implementation skills and grow through rigorous work on open-ended AI problems.
Algorithms are the foundation; systems are where ideas meet reality.
Competitive programming has given me a practical foundation in algorithms and problem solving, while my projects increasingly connect software, AI, and physical systems.
- Algorithmic reasoning through competitive programming
- Software systems built from interface to implementation
- Embedded work connecting sensing, control, and motion
02 / Research
Questions I am learning to pursue
These are current directions of exploration and study, not a list of established research contributions.
World Models
Representations that support prediction, reasoning, and decision-making about environments.
Embodied Intelligence
Connecting perception, planning, and control in agents that act in the physical world.
03 / Education
Academic foundation
Xiamen University Malaysia
Bachelor of Engineering in Software Engineering
04 / Evidence
Competitions & Distinctions
Evidence of practical experience in algorithms, implementation, teamwork, and mathematical modeling.
ICPC
Qualified for the 2026 ICPC Asia Pacific Championship in Taoyuan through the 2025 ICPC Asia Jakarta Regional Contest.
Team-based algorithmic problem solvingTHUPC
Bronze Medal, THUPC 2026.
Competitive programmingCodeforces
Reached a Codeforces rating of 1456 within the first several rated contests.
Algorithms and rapid problem solvingMCM/ICM
Contributed the core modeling and implementation framework to a 2026 MCM/ICM submission that received Honorable Mention.
Mathematical modeling and implementation
05 / Selected work
One focused product, built around a clear constraint
DisposableResume is my primary independent project: a small browser product that treats privacy as an architectural boundary.
Featured independent project
DisposableResume
A privacy-first, zero-retention résumé builder that runs entirely in the browser, with live preview and client-side PDF export.
Problem. Résumé tools frequently place sensitive personal data behind accounts, persistent storage, or server-side export pipelines.
No account, backend, database, analytics, or default persistent browser storage.
Additional engineering work
The complete projects page also covers IntelligentTransformCart, a collaborative embedded-systems prototype, and AutoPet, an independent local desktop-pet prototype.
06 / Independent learning
Coursework beyond the degree
A deliberate path through programming foundations, systems, machine learning, and language-model engineering.
Self-directed study using publicly available course materials; not formal enrollment or certification.
Completed
- UC Berkeley CS61ACompleted
- UC Berkeley CS61BCompleted
- UC Berkeley CS61CCompleted
- Stanford CS229Completed
In progress
- Stanford CS336 — Language Modeling from ScratchIn progress
- Carnegie Mellon University 11-868 — Large Language Model SystemsIn progress
07 / Contact
Open to rigorous, implementation-led research work
I am interested in undergraduate research opportunities and collaborations involving AI agents, world models, embodied intelligence, and related intelligent systems.