Initializing System...
Initializing System...
Published research and applied statistical analysis.
Ambiguous analytical questions converted into structured modelling work, using forecasting, simulation, and statistical methods to test assumptions and produce decision-ready evidence.
Modelling
Academic and industry projects requiring rigorous statistical analysis, predictive modelling, and simulation — from cryptocurrency market dynamics to logistics optimisation and healthcare operations improvement.
A combination of academic research (MSc/BSc thesis level) and applied industry projects. The research context required methodological rigour and peer-review-standard documentation; the industry projects required models that were explainable and actionable for non-technical stakeholders.
A portfolio of statistical modelling and simulation work demonstrating how ambiguous analytical questions can be converted into structured models, tested assumptions, and decision-ready evidence.
Multiple independent modelling pipelines: network analysis for the crypto research, a supervised learning pipeline for marketing propensity modelling, and discrete-event simulation for operational projects.
Applied the right methodology for each problem rather than forcing a single approach. In industry projects, focused on producing models that improved a measurable business outcome — not just high accuracy scores. Presented results in terms of business impact (marketing ROI, throughput improvement) rather than statistical metrics.
The methodological decisions were shaped by the nature of each problem. Social network dynamics in crypto markets are emergent and non-linear — agent-based simulation captured this better than traditional econometric models. For marketing propensity, prioritised recall over precision given the asymmetric cost of missing a convertible customer.