See how users chat with the AI assistant and how they use the MCP server.
An on-call engineer reviews their personally assigned alerts, inspects the most urgent one, accepts responsibility, and documents an initial investigation note.
An incident commander declares a customer-facing outage affecting multiple services, then updates the incident status as the root cause is identified.
A responder discovers an alert was routed to the wrong team, finds the correct escalation policy, and reroutes it to the database team.
A support lead manually creates a HIGH-priority alert for a customer-reported issue that monitoring didn't catch, wiring up the right source, escalation policy, and an extra responder.
An engineer working an accepted alert pulls in the DevOps on-call schedule for help, runs an automated restart action, and then resolves the alert with a documented fix.