I build neural network surrogates for granular material simulation and physics-informed ML systems and I'm increasingly focused on applied AI engineering for digital twin, predictive maintenance, and environmental intelligence. Currently exploring industry roles in ML/AI engineering, data science and computational modelling across Australia.
- 👨🎓 PhD in Engineering — Federation University Australia (collaboration with CSIRO, fully RTP-funded) Thesis: "Accelerated Surrogate Modelling of Granular Materials using Artificial Neural Networks"
- 👨🎓 Master of Engineering, Mechatronics (GPA: 6.2/7.0) — Australian National University
- 👨🎓 Bachelor of Engineering, Mechatronics — Ho Chi Minh City University of Technology, Vietnam
- Deep Learning for physics simulation
- Discrete Element Method (DEM) & Surrogate Modelling
- Physics-Informed Neural Networks
- Computational Modelling & Simulation
- Digital Twin & Predictive Maintenance
- Environmental Intelligence
| Project | Description |
|---|---|
| HoistDigiTwin (demo) | Digital twin of a mine hoist circuit end to end: a SimPy discrete-event simulation validated against a reference production year, an XGBoost/Gaussian process surrogate accurate to about 2% and under a millisecond per prediction, a remaining-life model proven on NASA's C-MAPSS data then applied to the twin's own winder, and a live React dashboard with MLflow-tracked experiments. |
| simple_notebooklm | A document-grounded RAG learning assistant inspired by NotebookLM — upload PDFs, get grounded Q&A, summaries, quizzes, and flashcards with source citations. |
| urbancool_melbourne (demo) | ML dashboard predicting urban heat vulnerability across Greater Melbourne from a Random Forest model (R² = 0.57): FastAPI backend, interactive SHAP explanations, and a live green-infrastructure what-if simulator. |
| eco_risk_au | Geospatial + LLM ecosystem climate-risk platform for Victoria, Australia: PostGIS ETL pipeline computing a composite climate risk score per vegetation patch, a LangGraph + Gemini RAG pipeline extracting structured risk data from real IUCN Red List of Ecosystems documents, and a React + Leaflet dashboard. |
🤖 Digital Twin & Simulation
📈 Predictive Maintenance & ML
🌍 Environmental & Geospatial Intelligence
Open to ML/AI engineering, data science, and computational modelling roles in Australia — feel free to reach out!