We develop explainable AI for high-stakes domains, from finance and telecommunications to cybersecurity and space.
AI for critical systems
Decision transformers, offline RL, and multi-agent planning that adapt to real-world constraints.
Intrusion detection, rule selection, and privacy-preserving design for critical systems.
XAI for audit, certification, and human-over-the-loop assurance.
ML + operations research for robust policies under uncertainty.
Work across cyber, space, finance, and health
Autonomous AI agents for proactive cyber defence across a 12-partner European consortium.
Decision Transformer-based Suricata rule management for adaptive intrusion detection without downtime.
FNR-funded research on explainable clustering, classification, and resampling for financial services.
Retrieval-grounded risk analysis for portfolio managers with cited, explainable market insights.
Multi-objective RL and operations research for adaptive 5G network self-optimisation.
Zero Trust-oriented security research for connected and autonomous vehicle systems.
Security and privacy architectures for genomic, biomedical, and healthcare data environments.
GAN-based cloud removal research underpinning remote sensing and geospatial AI expertise.
We work with partners in the NC3, Luxembourg House of Cybersecurity (LHC), LIST, and government agencies. Our work contributes to EU programmes including Digital Europe and defence innovation calls.