🔭 What I do Building AI products, mentoring early-career data professionals, and consulting on data science & AI strategy.

🌱 Currently exploring Claude Code, Codex, MCP, and agentic workflows.

👯 Looking to collaborate on Real-world DS/AI projects, open‑source tools, and knowledge‑sharing content (blogs, talks, workshops).

💬 Ask me about Data science career growth, LLM‑powered applications, or turning messy data into decisions.

📫 Let's connect X/Twitter · LinkedIn

⚡ Fun fact I'm currently mentoring 10+ rising data professionals – mostly over Google Meet and an unhealthy amount of coffee ☕

• lux-tts — Voice cloning in Google Colab from a short audio sample, no GPU required
• persona — Synthetic persona generation and interview skill for Claude Code
• melbourne-property-intelligence — LLM-powered property market intelligence with RAG, FastAPI, and Streamlit

• how-to-spell — Web app for kids to learn spelling by speaking
• money-game — Teach young kids about coins

• autokeras-text-classification — Spam detection and product review analysis with zero architecture decisions
• pycaret — Predicting employee attrition — comparing 10+ classifiers with SHAP interpretability
• lazypredict — Benchmarking every ML algorithm in one line of code
• prophet-time-series — Airline demand forecasting with seasonal decomposition and confidence intervals
• regression-using-tensorflow — Predicting fuel efficiency with a TensorFlow/Keras neural network
• review-data-using-SVC — Sentiment classification of restaurant reviews — full NLP pipeline
• dython — Measuring associations between categorical variables with Theil's U heatmaps
• google-cloud-vision — Image labelling, face detection, and visual search via Cloud Vision API

• ml-with-r — Predicting diabetes with GLM — full evaluation with accuracy, recall, precision, F1, AUC
• kaggle-predict-house-prices — Kaggle competition — XGBoost on 80 features with modelStudio explainability
• predict-marketing-response-with-xgboost — XGBoost for marketing campaign response — customer segmentation
• tidymodels — Decision Tree vs Random Forest vs XGBoost — progressive comparison with tuning
• model-studio — Comparing model explainability across algorithms with interactive DALEX dashboards
• linear-regression-in-r — Simple vs multiple regression — Facebook has 4x the per-dollar return of YouTube
• logistic-regression-in-r — Three datasets, coefficient interpretation, and probability math
• ggtext — Rich text in ggplot2 — colour-highlighted titles and styled axis labels

• image-tinder — Swipe through photos like Tinder — keep the good ones, ditch the rest
• slaps — Two-player face-to-face card game for mobile browsers
• noise-monitor — Real-time noise level monitor using Web Audio API

• statistical-learning-data-mining — Airbnb Sydney price prediction using statistical learning techniques
• proj-auction-results — Scraping and visualising Melbourne auction results with Leaflet maps