Data Science · Business Analytics · Data Analytics
Applied analytics, decision support, and business-facing data projects built around real commercial, operational, finance, and healthcare use cases.
About me
I build business-facing analytics that turn operational, commercial, and financial data into clearer decisions, structured visibility, and more useful reporting.
My background combines recent training in Data Science and Business Analytics with a strong interest in solving real business problems through structured analytical work.
I am especially interested in projects where data is not only explored, but translated into tools, decision-support systems, and outputs that companies can actually use.
Selected work
Explore the portfolio by role positioning. Each case is designed to show how I use data to clarify problems, diagnose drivers, and support management action.
A Business Analyst case showing how fragmented internal service request intake can be translated into structured requirements, validation rules, workflow controls, KPI reporting, and prioritised management action.
Operational analytics case for a diagnostic lab network, focused on SLA reliability, workflow bottlenecks, rework impact, and site-level quality action prioritization.
A business-facing analytics case focused on discount leakage, returns, weak price realization, and margin pressure in an omnichannel retail business.
Driver-based FP&A planning and Power BI reporting case showing how stable demand can still translate into weaker profitability when discounting, returns, and cost-to-serve pressure build below revenue.
A management-facing controlling case built around a harder profitability question: which products still create real value once setup effort, fragmentation, fulfillment burden, service support, and scarce capacity are taken seriously?
Finance-focused analytics project designed to demonstrate source-to-target logic, variance analysis, reconciliation controls, and stakeholder-facing reporting.
A product analytics case built around a real marketplace question: did a new discovery treatment improve user progression enough to justify selective rollout?
A full analytical system turning fragmented commercial data into structured visibility, segmentation, recommendation logic, and business-facing outputs.
A production-style AI reliability case built around a harder question than demo quality: when can a claims knowledge assistant be trusted enough to scale, and how should unsafe behavior be detected, reviewed, and rolled back after release?
A grounded GenAI workflow built around a realistic internal operations problem: helping teams get faster, more reliable answers from scattered policies, procedures, approvals, and support documents.
Flagship project
This master’s project was built around a real business need: turning scattered commercial data into a practical system for visibility, analysis, and decision-making.
Designed around an actual business problem rather than a purely academic exercise.
From raw data and validation to dashboarding, analytical outputs, and recommendations.
Created to improve visibility, support action, and make commercial decisions clearer.
The starting point
Sales, customers, and product data were spread across multiple operational sources. The challenge was not only to analyse the data, but to transform it into something structured, reliable, and actually useful for business decisions.
How the system works
The project combined data ingestion, cleaning, quality control, harmonisation, and analytical modelling into a structured workflow capable of supporting reporting, segmentation, recommendation logic, and decision-ready outputs.
Visual proof
Each dashboard view supports a different layer of decision-making: overview, diagnosis, customer understanding, and next-best commercial action.
Business problem
The project addressed a common SME problem: having commercial data available, but lacking a structured system to convert it into useful visibility and decisions.
Technical execution
It required data engineering, validation logic, analytical modelling, dashboard design, and the creation of outputs that could actually be reused and scaled.
Business value
The final objective was not just insight generation, but helping a business see more clearly, prioritise better, and act with more confidence.
A high-level commercial view of revenue, customers, repeat purchasing, ticket size, and channel distribution.