Data Analyst with a machine learning focus, based in Sydney. I spent 3.5 years running operations and P&L for a community sports non-profit, then made a deliberate pivot into data: 99% average in the HyperionDev/UNSW Data Science Bootcamp, Microsoft PL-300 certified, and a portfolio of deployed projects below.
I build things to understand them. Everything here runs.
| Project | What it does | Stack | Live |
|---|---|---|---|
| OwnYourShip | CLI that quizzes you on your own LLM-generated code - you shouldn't ship what you can't explain | Python, FastAPI, SQLite, Anthropic API | - |
| House Price Predictor | XGBoost model, 95.27% R² on Sydney property data, 67% error reduction vs baseline | Python, XGBoost, Streamlit | Live |
| Neural Net from Scratch | Full MLP in pure NumPy - 97.28% on MNIST, 2.2x faster than sklearn | Python, NumPy | - |
| Transformer Attention Visualiser | Interactive app showing attention patterns forming in real time - 500+ users | Python, NumPy, Streamlit | Live |
| RL Foundations: Sutton & Barto | Every algorithm from Chapters 1-13, tabular (NumPy) + deep (PyTorch) | Python, PyTorch, Gymnasium | Live |
| Power BI Sales Analysis | Star schema + 15 DAX measures; surfaced $8.9M recovery opportunity | Power BI, DAX | - |
Python · SQL (PostgreSQL, SQLite) · Pandas · NumPy · scikit-learn · XGBoost · PyTorch · TensorFlow · Power BI (PL-300) · Streamlit · Git · Docker
