Portfolio
Data science and econometrics projects combining economics, statistics and modern analytics to solve real-world problems.
Built a global ML index and interactive web app predicting migration and displacement pressure using climate, conflict and food security data at subnational level, deployed with Python and web technologies.
Developed Python-based models to predict English Premier League match and season outcomes using historical performance, xG statistics and team features, comparing classification and regression approaches.
Designed an NLP pipeline in Python to detect online rumours using text preprocessing, feature engineering and ML classifiers, improving detection accuracy on labelled social media datasets.
Led data science work using Python, R, SQL and Power BI to analyse acquisition funnels, segment donors and build churn-prediction models. Reduced churn by 40% across OXFAM, UNICEF and the Red Cross.
Built interactive Power BI dashboards exploring US population, internal migration and demographic trends, enabling stakeholders to explore regional patterns and changes over time.
Analysed global COVID-19 case trends and mobility indicators using Python, building time-series forecasting models to study transmission dynamics and short-term projections.