SentinelAI is JIL's AI-powered fleet monitoring system that continuously inspects all validator nodes across the network. It monitors heartbeat metrics, evaluates threat scores, and executes automated response actions based on configurable policies. The system runs fleet-wide inspection cycles with anti-loop protection limiting cycles to 3 per 2-hour window.
Managing a distributed validator fleet across 13 jurisdictions requires continuous automated monitoring. Manual monitoring cannot scale to the volume of metrics generated by 10+ validators each running dozens of services. Delayed detection of validator issues can compromise network security and settlement integrity.
SentinelAI provides real-time fleet intelligence. Each validator sends regular heartbeats containing health metrics, service status, and resource utilization. SentinelAI analyzes these signals, computes threat scores, and triggers automated actions - from alerts to service restarts to validator isolation. The inspection rules and auto-action policies are configurable per validator and per threat level.
SentinelAI monitors validator heartbeats, service health, resource utilization, network latency, consensus participation, and security posture. It computes threat scores and triggers automated responses based on configurable policies.
Anti-loop protection limits fleet-wide cycles to 3 per 2-hour window. Threat scoring uses multiple signals and configurable thresholds. Automated actions escalate from alerts to restarts to isolation, preventing overreaction to transient issues.