I specialise in predictive modelling, operational intelligence, and decision-support analytics — bridging technical depth with business context across housing, healthcare, and e-commerce.
My toolkit spans the full analytics pipeline — from raw data engineering and modelling through to executive-ready dashboards and automation. I choose tools for the problem, not the other way around.
Applied Random Forest, KNN, and Decision Tree models to the UCI Adult Dataset to predict income levels. Random Forest achieved ~85.7% accuracy with strong ROC-AUC performance, also implemented in Azure ML Designer.
Forecasted daily Wikipedia article views using ARIMA, LSTM, XGBoost, and Prophet. LSTM and Prophet demonstrated robust performance capturing complex seasonality and trends.
Applied K-Means and Hierarchical Clustering to UCI Online Retail data. Used WCSS Elbow and Silhouette Scoring to identify four distinct purchasing behaviour profiles.
Analysed Netflix content using VADER sentiment scoring, N-gram analysis, topic modelling, and word clouds to uncover emotional trends across genres and top actors.
Analysed global health systems 2010–2021 using R, examining the interplay between demographics, fiscal allocation, mortality, and equity across diverse socioeconomic contexts.
Diagnosed SME churn drivers for a European energy utility, framing price sensitivity as a data science problem and designing a predictive model to target a 20% discount strategy.
End-to-end data cleaning pipeline using SQL: addressed incomplete data, duplicate rows, address splitting, and format standardisation to maximise downstream analytical accuracy.
Interactive sales dashboard built in Looker Studio surfacing revenue, product performance, and coupon insights from Google Sheets data for a mental health tech organisation.
Whether you're looking for a data scientist to join your team, a consultant to drive an analytics project, or just want to talk data — I'd love to hear from you.