As a Data Science Intern you will collaborate to develop top‑notch data science products to fight financial economic crimes such as money laundering, fraud and financing of terrorism. During this internship, you will contribute to one of several advanced machine learning research projects within our team, working alongside experienced data scientists for guidance, feedback, and collaboration.
At the start of the internship, you will choose (in alignment with the team) to work on one of the following topics:
Project 1: Training machine learning models for fairness and coverage
Explore how model training and decision-making can be adapted to balance predictive performance with fairness across groups and coverage across client segments, using multi-objective optimization techniques.
Project 2: Stress testing machine learning models in the Financial & Economic Crime domain
Develop methods to systematically evaluate model behavior in underrepresented data dimensions (e.g. tails of distributions) by perturbing input data, helping uncover blind spots and improve robustness. As an additional objective, the project will explore the use of selected synthetic scenarios as part of the training or fine‑tuning process.
Project 3: Quantifying robustness and uncertainty in model scores
Investigate techniques to estimate and incorporate prediction uncertainty (e.g., confidence intervals or distributions) into decision-making to improve reliability and reduce spurious alerts.
