Investigation console
AI Pilot
Plan retrieval across your monitoring sources, build an evidence package, then reason over it. Every conclusion carries severity, confidence, evidence and a recommended action.
This session
Workspace history
Past investigations are stored in this browser workspace only.
Active contextInfrastructureCurrent Month
Situation
Root cause
Context
no enabled monitoring sourceno AI provider selected
No active investigation
Ask a question, or start an investigation directly from an incident or dashboard card. The pilot selects the relevant data sources, retrieves normalized telemetry and builds an evidence package before answering.
Evidence
Everything the answer is grounded in
Evidence collected for an investigation appears here: sources queried, entity state, metrics, timeline and log lines.