Caffeinated coder

AEGIS (Autonomous Evidence-based Graph Intelligence System) is an AI-powered financial fraud intelligence and investigation platform designed to identify and explain suspicious activity across interconnected financial transactions. The system combines deterministic transaction detection, graph intelligence, evidence validation, and Meta AI-powered investigation assistance. AEGIS identifies patterns such as circular transfers, rapid fund dispersal, and layered transaction chains. Its graph engine reconstructs transaction relationships, identifies intermediary accounts, and provides structural insights into financial activity. A deterministic evidence validation layer checks detected patterns against underlying transaction records, distinguishing verified observations from hypotheses and missing evidence. Meta AI models power the investigation intelligence pipeline, generating evidence-grounded summaries, explaining suspicious patterns, and providing advisory recommendations. Analysts can then review findings, inspect supporting evidence, and manage structured investigation cases. AEGIS uses synthetic financial data and is designed as a human-in-the-loop decision-support system. It does not autonomously block transactions or make final compliance decisions.