Research
FraudGuard AI
Detecting fraud from transactional and relational signal.
FraudGuard AI combines transaction intelligence, graph intelligence, and anti-money-laundering (AML) signals, with explainability and human validation, in a continuous-improvement loop.
Status
Published
Authors
Karim Bettaieb, Afef Kacem Echi, Houcem Hammami
Publication
International Journal of Data Science and Analysis · 2026 · DOI 10.11648/j.ijdsa.20261205.11
Abstract
This study introduces a coverage-aware and protocol-aware evaluation of account-risk enrichment, graph neural networks, tabular gradient boosting, and graph-tabular stacking for anti-money laundering and fraud detection.
Verifiable external publication. Research results are not a promise of customer performance.