Four years of data science and analytics

Part-time from September 2025, full-time since May 2026.

• Analyse store-level sales, transaction and labour data in SQL and Python to find revenue trends, performance gaps and labour-efficiency opportunities across restaurant operations.
• Built Power BI dashboards and KPI reporting on revenue, transactions, average transaction value, labour cost %, sales per labour hour and store productivity.
• Use statistical analysis and difference-in-differences to evaluate operational changes, comparing before and after performance across treatment and comparison stores to measure incremental impact.

• Developed a Power BI analytics platform for a US confectionery portfolio worth more than $1B, improving profitable revenue management.
• Built marketing mix models to quantify the effectiveness of marketing and trade promotions and guide investment decisions.
• Processed and analysed more than three years of sales, retail and promotional records in SQL, Python, PySpark and R for predictive modelling and commercial analytics.
• Designed trade promotion optimisation that improved trade-spend efficiency by 8% and supported profit-maximising promotional plans.
• Delivered analysis of promotional ROI, sales attribution and revenue drivers to support commercial decisions.
• Built and deployed analytics workflows on Azure Databricks, Data Factory, Data Lake, Azure SQL and Azure ML.
• Mentored junior team members through code development and reviews, and delivered with cross-functional stakeholders using Agile, CI/CD pipelines and Azure DevOps.

• Developed an RFM-based customer risk segmentation for more than 200,000 customers in SQL and Python, grouping debt profiles into high, medium and low risk. Rolled out through an A/B test, the targeted recovery strategy raised recovery rates by 14% and cut average recovery time by 12 days.
• Developed executive Power BI dashboards on customer segments, debt ageing, recovery performance and portfolio exposure for senior leadership.

• Processed more than one million insurance claims and policy records in SQL to identify operational bottlenecks and the main drivers of SLA breaches, supporting process improvement.
• Developed Tableau dashboards and analytics for business reporting, including the data transformation and feature engineering behind them.