The Request Agent node
The Request Agent node is the handoff mechanism in workflows. When the conversation reaches this node, the bot sends a customizable message to the visitor and then creates an unassigned conversation in your inbox. An agent picks it up from there.
Adding a Request Agent node
- In the flow editor, click the + handle on the node before the handoff point.
- Select Request Agent.
- Edit the pre-handoff message — this is what the visitor reads before the agent joins. For example: "Connecting you to our team now. An agent will be with you shortly."
- Connect the node's output if needed (some flows continue with a Reply Time node before Request Agent, to set expectations).
What the agent sees
When an agent opens the conversation in the inbox:
- The full workflow conversation history is visible — every message, button click, and data collected.
- Any visitor data collected during the workflow (name, email, custom fields) is shown in the conversation sidebar.
- The agent continues the conversation from the same thread — visitors don't need to repeat themselves.
Assignment rules
By default, workflow-escalated conversations land as unassigned and are available to all agents. If you have assignment rules configured (for example, route conversations with the topic "billing" to the billing team), those rules apply to workflow handoffs too. Set up assignment rules in Settings → Assignment rules.
What happens when no agents are online
If no agents are available when the visitor reaches the Request Agent node, the conversation is queued as unassigned. The visitor receives your offline message (configured in Settings → Business hours). The conversation appears in the inbox as soon as an agent comes online.
If you want to set visitor expectations about wait time, add a Reply Time node just before the Request Agent node.
Handing off to Vex AI instead
If your goal is to route from a workflow to AI rather than to a human, use an AI Bot node instead of Request Agent. Vex takes over and answers questions using your knowledge sources. For a comparison of when to use which, see Which chatbot should I use — Vex AI or workflows?
Next step
Learn how multiple workflows run in parallel without conflicting: Run multiple workflows at the same time.