Julien Mesrine

Leading and structuring a data science community (10+ analysts and data scientists): sharing best practices, project experience feedback, methodological support, promotion of machine learning to business units. Designing internal tools facilitating the creation, evaluation, and deployment of models (feature selection, drift monitoring, performance tracking), in collaboration with the development team, to make data teams more autonomous. Carrying out impact projects, including: Real-time fraud prevention: reduced fraud cost by 3x. Incoming call volume forecasting: optimized customer service staffing. Marketing mix model: decision support tool for media budget allocation.

Developing fraud detection algorithms: business rules, statistical models, continuous optimization of existing systems. Designing monitoring dashboards (fraud KPIs, algorithm performance) for business and management teams. Producing regulatory reports (OSCAMPS – Bank of France).

Developing a series of fraud prevention algorithms in SQL within the Risk department.