At KHAN Bank, I administer WAF and DDoS protection for internet-facing banking services and take part in attack analysis and response. I assess externally facing systems against the OWASP Top 10, contribute to fraud detection and investigations, and support PCI DSS compliance.
My research connects banking fraud techniques with risks in AI-mediated environments. I authored an open-access framework on cognitive harm and built attack-kg-agent, an open-source prototype that validates and repairs LLM outputs against a MITRE ATT&CK knowledge graph; on 300 SigmaHQ rules, graph repair raised top-1 accuracy from 41% to 60% without extra model calls.
