Data scientist (MSc, Distinction) who found a passion for analytical problem-solving during a successful international finance career. Spent 8+ years in progressively senior roles managing strategic analysis for billion-dollar operations, where the most rewarding work lay in spotting hidden patterns, asking better questions, and building compelling data stories that changed minds.
That investigative streak led to formal study in Data Science & AI, with a thesis on modern generative approaches and foundation models for disaster image classification, alongside building robust skills in machine learning, computer vision, and statistical modelling.
Now seeking research-oriented roles across diverse domains - from crisis informatics to environmental analysis to social research - and particularly drawn to interdisciplinary projects where data science meets real-world impact. Happy to take hands-on, early-career roles where I can pair proven analytical leadership with the chance to work on complex, meaningful challenges.
- Benchmarked against published CNN models trained on MEDIC dataset (natural disasters and humanitarian response imagery)
- Compared two approaches: synthetic data augmentation (generating 10K additional training images) versus direct classification using vision models
- Vision models significantly outperformed both CNN benchmarks and synthetic data approaches
- Technical implementation: Python, PyTorch, TensorFlow, OpenAI APIs, advanced prompt engineering
- Worked across Marketing, R&D, and regional leadership to align commercial priorities with operational feasibility, building consensus and securing buy-in for complex strategic plans
- Designed and implemented an Anaplan tool to automate financial modelling for new launches, incorporating R&D cost inputs and automatically performing cannibalization and pricing analysis from live product data
- Initiated and co-designed a Europe-wide Data Analytics Training Programme for finance teams, evangelising and teaching low-code tools (Power BI, KNIME) to automate workflows and perform advanced analysis—an initiative that finalised the decision to pivot into pure data science
- Responsible for tooling to run the entire forecasting process entirely offline initially, then co-designing replacement forecasting system in Oracle
- Trained and supported global teams in adopting new forecasting processes and tools, ensuring consistency and reliability during corporate transition
- Navigated shift from P&G's accuracy-focused forecasting culture to Coty's target-driven environment, protecting data integrity despite wider organisational pressures
- Consolidated multi-region SKU-level data into a unified forecasting platform, enabling transparent performance tracking and informed decision-making
- Developed anomaly detection models to identify irregular distribution patterns, ensuring compliance and market integrity in conjunction with legal team
- Responsible for data-driven distributor negotiation strategy, often directly involved in distributor negotiations along with Sales team for larger distributors
- Developed forecasting frameworks and analytical models to support strategic business planning
- External audit experience