Incident Correlation
Automatically correlate events across providers, services, and time windows to surface the root cause of complex incidents.
Features
Cross-Provider Correlation
Detect when issues span multiple LLM providers — e.g., a latency spike on OpenAI coinciding with increased error rates on Anthropic fallback. Events are grouped into a single incident.
Timeline Reconstruction
Unified timeline showing all related events in chronological order — alerts, security events, provider status changes, configuration changes, and user actions.
AI Analysis
The correlation engine uses AI to identify probable root causes: "OpenAI rate limit hit at 14:02 → fallback to Anthropic → Anthropic circuit breaker opened at 14:05 → all requests failing."
Dependency Mapping
Visualize how incidents propagate through your AI stack. See which downstream services are affected by an upstream provider outage.