Finesse Lab wordmark

Finesse Lab is part of the Socially Intelligent Artificial Systems (SIAS) group within the Informatics Institute (IvI) at the University of Amsterdam, with strong ties to the Institute for Logic, Language and Computation (ILLC) and the AI4FinTech initiative. In Finesse Lab, we pursue foundational neurosymbolic AI research for safety and explainability inspired by industry-relevant use cases in financial services. Our work is motivated by the view that high-stakes application domains are not merely places to deploy AI, but rigorous environments in which new methods for responsible, trustworthy, and reasoning-aware AI can be developed and tested.

Finance is especially compelling in this respect because AI systems increasingly shape decisions around credit, fraud detection, compliance, insurance, payment systems, risk, and policy design. These are settings in which predictive performance alone is not enough. Systems must also support transparency, accountability, meaningful human oversight, and clear reasoning about why decisions are made. In Finesse Lab, our work is therefore finance-inspired rather than finance-exclusive: behind each project lies a finance-relevant scenario that sharpens the scientific questions around safety, explanation, reasoning, forecasting, and control.

Methodologically, the lab focuses on neurosymbolic systems that integrate data-driven learning with logic, rules, causality, and domain knowledge. This includes work on multiagent systems, mechanistic interpretability, tabular foundation models, and forecasting in settings where explanation and reliable reasoning matter. More broadly, our research follows a for people, by people, with people philosophy: we begin from stakeholder needs such as fairness, transparency, accountability, and recourse; we incorporate expert knowledge directly into models through constraints and structured reasoning; and we design AI systems that support analysts, auditors, compliance professionals, and citizens rather than replacing them with opaque automation. A central part of the lab’s mission is to engage industry collaborators from the finance and fintech domain so that foundational research can also lead to real-world impact.