Escalation Insights helps you understand the escalation patterns in your AI + human support workflow. It shows when your AI agent hands off conversations to human agents, the reasons behind those escalations, and trends over time.
This feature is coming soon and is not yet available. The page will be updated when it launches.
What to expect
When Escalation Insights launches, you’ll be able to see:
- Escalation volume — how many conversations were escalated in a given period
- Escalation rate — what percentage of total conversations required human intervention
- Top escalation reasons — the most common triggers for escalation (e.g. topic the agent couldn’t handle, user request, specific keywords)
- Escalation by channel — which channels generate the most escalations
- Time to escalation — how long into a conversation before escalation typically occurs
- Resolution after escalation — whether escalated conversations were ultimately resolved by the human agent
Why this matters
Escalation Insights is a powerful feedback loop for improving your AI agent. High escalation rates on specific topics mean:
- Your agent may lack knowledge in that area → update the Knowledge section
- The agent’s skills for that scenario may need tuning → review Skills settings
- A Workflow may help structure conversations in that topic more effectively
Use this view to systematically reduce escalations over time by addressing their root causes. 
