MSc Student · AI Systems / ML Engineering
GPU Scheduling · Agent Workflows · Computer Vision
Building reproducible AI systems across GPU scheduling, agent workflows and computer vision.
Python PyTorch Linux CUDA FastAPI Git
Interested in AI Infrastructure · Agent Engineering · GPU / Inference · Developer Tools. Open to remote AI/ML internships and technical collaborations.
| Project | What I built |
|---|---|
| Forecast-Aware GPU Scheduler | Future-workflow prediction for online GPU scheduling: node-level event simulation, memory and model-cache constraints, scheduling baselines, and reproducible benchmarks. |
| Research Agent OS | An installable core for long-running AI-assisted research: persistent project state, separate execution and review roles, experiment gates, and context handoffs. |
| Real-Time Video Analytics | YOLO + ByteTrack → counting / ROI / dwell events → FastAPI dashboard, with MP4 and basic RTSP support. The repository documents 46 passing tests. |
| Solar Filament Segmentation | PyTorch research on low-contrast solar filament instances: tiled U-Net, auxiliary heads, frozen splits, and auditable experiment records. |
More: ChatGPT Web Bridge — developer tooling with an attach-only CDP bridge, skill / MCP integration, and explicit execution boundaries. YOLO Optimization Portfolio — dataset audit, FP/FN diagnosis, and controlled evaluation.
South China University of Technology — Master's student in Mechanical Engineering, 2025–present; B.Eng. in Vehicle Engineering.
华南理工大学机械工程硕士研究生，关注 AI 系统与机器学习工程，主要开展 GPU 调度、Agent 工作流与计算机视觉项目。重视可复现的研究和完整的工程实现，寻求远程 AI/ML 实习与技术合作。