Student session
From idea to interface in an afternoon
By Lars Janssen, Tom Peters and Sam Neefjes (PWC)
In this training, we’ll build small web applications together, live. Using simple front-end tools, we’ll show how PwC’s actuarial team turns econometric and actuarial ideas into clickable prototypes. You’ll design and build your own mini app with guidance, while we explain each step and link it to real actuarial and consulting work. The goal is to show how rapid prototyping makes complex insights tangible, easy to test, and clear to communicate.
Interactive I
GenAI assisted model validation process
By Maarten van der Maarel and Bart Rikkert (Deloitte)
In this session, Deloitte and Aegon will present their joint exploration of a GenAI implementation in the Model Validation process, including challenges, benefits and lessons learned.
Interactive II
From Classical Machine Learning to Agentic AI: Scaling a Data-Driven Insurer
By Harm Mulder and Emiel Bangert (Top Advisory)
Join us for From Classical Machine Learning to Agentic AI: Scaling a Data‑Driven Insurer, showcasing how a startup insurer with a fully self‑developed reporting and data backbone has been built. Explore how AI is embedded throughout to enable a truly modern insurance operation."
Interactive III
How to prevent discrimination in AI Pricing? A Fair Pricing Game
By Rijk Mercuur (Hogeschool Utrecht)
This interactive session explores how actuaries face the risk of indirect discrimination, as proxies such as postal codes in AI models can create significant and hard-to-detect disparities, leading to increasing legal and reputational risks under the AWGB and the AI Act. While fairness metrics make these effects technically measurable, they differ substantially in definition and require difficult trade-offs between ethics, explainability, and profitability. In a serious game, participants take on the role of an insurance manager and experience how choices in data selection, fairness metrics, and model adjustments affect both fairness outcomes and financial performance.