Nexus AML is a high-performance, graph-based AI framework designed to detect complex money laundering patterns in real-time. By modeling financial transactions as directed temporal graphs, the system identifies sophisticated criminal networks—such as circular layering and fan-out schemes—that traditional rule-based tools often miss.
Built for massive scale, the platform utilizes a GraphSAGE neural network architecture to process approximately 17,900 transactions per second. It achieved an ROC-AUC of 0.912 and an F1-score of 0.846 during testing on the IBM AMLSim dataset of 179 million transactions. To ensure regulatory compliance and trust, Nexus AML includes an explainable AI (XAI) pipeline that provides investigators with clear visual evidence and structural justifications for every flagged account.
Log in or sign up for Devpost to join the conversation.