Modern banking platforms must deliver uninterrupted service despite rising architectural complexity, strict regulatory expectations, growing cyber risk, and constant pressure for faster change. Traditional resilience practices often rely on static controls, post-incident reviews, and manual analysis, which are too slow for distributed payment and core banking systems operating in real time.

This keynote presents a practical framework for resilience engineering in banking platforms using chaos engineering and AI feedback loops. The central idea is simple: controlled fault injection becomes far more valuable when paired with machine learning models that continuously learn from telemetry, dependency behavior, deployment patterns, latency signals, and security events. Instead of treating chaos experiments as isolated tests, this approach turns them into an intelligent feedback system for production resilience.

The session will show how chaos experiments can be used to simulate service degradation, dependency failure, tail-latency amplification, and control-plane disruptions across banking workloads. AI-driven analysis then correlates those signals to identify weak points, predict abnormal behavior earlier, estimate blast radius, and prioritize remediation actions. By integrating observability, security telemetry, and intelligent decisioning, teams can move from reactive firefighting to proactive resilience improvement.

Attendees will gain a blueprint for combining chaos engineering with applied intelligence to improve fault tolerance, reduce mean time to detect, strengthen security posture, and build more adaptive, trustworthy banking systems.

https://scrs.in/conference/aic2026/speaker/talk/22869