How to hand off an AI chatbot to human support
Create a dependable escalation path and practice the full agent takeover workflow.
Human handoff works only when both sides are configured: the chatbot must request escalation at the right moment, and an available agent must see, claim and end the session.
The timeout prevents visitors from remaining in silent human mode when nobody responds. Configure it together with staffing and notifications.
What you will have at the end
- An enabled Live Chat tool
- A defined escalation and timeout policy
- A completed test takeover in the Live Chat dashboard
Before you start
- Standard plan or higher
- At least one team member with Live Chat permission
- A monitored notification channel
A handoff is a state transition
The AI handles the conversation until a visitor or escalation rule requests a human. An agent claims the session, writes directly to the visitor and explicitly ends it so the bot can resume later.
If nobody claims the request within the configured timeout, the bot resumes automatically instead of leaving the visitor stranded.
01–05
Set it up step by step
Enable Live Chat and set the timeout
The full tool configuration stays visible in one light-mode frame.
Open the chatbot → Tools → Live Chat, enable the tool and choose a wait time that matches real staffing. The captured configuration keeps the activation, availability and timeout controls visible together.
Request a human from the visitor chat
Verify the visitor enters the waiting state.
In a fresh visitor chat, explicitly ask for a human and provide the tutorial contact details. The real chat shows the handoff acknowledgement and waiting state; this confirms that the request left normal bot mode.
Open the request and review its history
The agent sees the complete context before replying.
Open Dashboard → Live Chats and select the waiting request. Review the visitor identity and the complete conversation history before taking over; the screenshot shows both the request and its context in the dashboard.
Take over and send the agent reply
Prove the human message was sent in the active session.
Select Take Over, write a concise agent response and send it. The captured state shows the active human session and the delivered reply, not merely the empty Live Chat interface.
End the human session
Return control deliberately so the bot can answer again.
Select End Chat in the dashboard and confirm that the session is closed. Then send another visitor message; the practical example below shows the bot resuming with a grounded NORDSTERN-42 answer after the human session ended.
Example & result
See the practical test and its result
Every tutorial includes a fixed input, the expected outcome and a transparent record of what was actually verified locally.
Practical example: How to hand off an AI chatbot to human support
This exact scenario was completed with the temporary tutorial account.
Exact test input
Please connect me to a human support agent. My name is Alex Example and my email is alex@example.com.
Expected result
The visitor enters waiting mode, the agent sees the complete history, replies and ends the session; the bot then resumes.
What was actually verified
The complete flow was executed: bot answer → handoff → agent reply → session end → grounded bot answer.
Tips & tricks
Make the setup reliable
Test with realistic examples, record your baseline and change one setting at a time. That makes real improvements visible.
Show honest availability
Do not promise a person when nobody is staffed. Outside hours, collect a callback or lead instead.
Prepare concise templates
Templates for greeting, investigation and closure reduce response time while agents keep full context.
When something does not work
Troubleshooting
Check status, permissions and test data systematically before changing the model or prompt.
The expected option is missing
Confirm the account plan, feature permissions and selected chatbot. Paid or beta features can be hidden when prerequisites are not met.
The test result is inconsistent
Reset the test conversation, keep the input identical and change one setting at a time so the cause remains measurable.
Ready for a production-style test
Measure request volume, claim time and abandoned handoffs. Adjust staffing, timeout and escalation rules together rather than treating them as separate settings.