Aspiring AI + Materials Science Researcher · Full-Stack AI Engineer · Astrophotography Enthusiast
I'm working at the intersection of AI and Materials Science, with a focus on:
- Graph Neural Networks for molecular property prediction & materials design
- Machine Learning Interatomic Potentials (MACE-MP-0, CHGNet, MatterSim)
- Computer Vision for Human-Robot Collaboration — action segmentation in bimanual assembly tasks
- Generative Inverse Design for battery materials (solid-state electrolytes, sodium-ion, organic electrodes)
🎯 Goal: PhD in AI for Science (targeting MBZUAI / KAUST, 2027 intake)
Currently collaborating with researchers at the University of Auckland (Computer Vision / HRC) and exploring AI for Science directions in battery materials intelligent design.
ML / AI
Full-Stack & Automation
AI Engineering & Infra
Hardware
RTX 4060 Laptop · RTX 5060 Desktop · 4×A100 80GB Cluster (University)
| Project | Description | Stack |
|---|---|---|
| 🏭 PG120 Smart Manufacturing Platform | Production line management system for precision planetary gearboxes — 54 API endpoints, 5 core modules + 4 AI modules. Built for Huawei Cloud xDM-F MiniApp Competition. | Vue3 · Express · Tongyi Qianwen · xDM-F |
| 🧪 Materials Science LLM Pipeline | Baseline evaluation → distillation data generation → LoRA fine-tuning loop for domain-specific LLM on 4×A100 cluster. | DeepSeek-R1 · vLLM · LLaMA-Factory |
| 📰 FinAI Daily (财智日报) | Fully automated content pipeline: financial/AI news scraping → podcast scripts, WeChat articles, Xiaohongshu posts. | n8n · Claude API · OpenClaw |
| 🎬 Intelligent Cockpit Micro-Course | Directed production of AI-powered micro-course series for Chun Tai Technology (2025.08–2026.01). | AI Video · Content Production |
| 🔬 GNN + Action Segmentation (in progress) | Research on bimanual assembly action segmentation using GNN + dilated causal convolution, based on DuHa-v2/HA-ViD dataset. | PyTorch · PyG · Conv1d |
"Build what matters. Measure what counts. Ship what works."
