Hand off to a live agent
With Power Virtual Agents, you can hand off conversations to live agents seamlessly and contextually.
When you hand off a conversation, you share the full history of the conversation (the context) and all user-defined variables. With this context, a live agent that uses a connected engagement hub can be: notified that a conversation requires a live agent, see the context of the prior conversation, and resume the conversation.
For more information about how to configure hand-off with Omnichannel for Customer Service, see the Configure hand-off to Omnichannel for Customer Service topic.
You can choose to escalate a bot conversation without linking to an engagement hub:
- In the Authoring canvas for the topic you want to add an escalation option to, at the end of the topic, select Add node (+) to add a new node.
- Select Go to another topic and then Escalate.
Escalate is a system topic that, by default, provides a simple message to a user if they ask for a human agent.
You can edit the topic to include a simple URL to a support website or ticketing system, or to include instructions for emailing or contacting support.
- Learn more about what you can do with Power Virtual Agents.
- A bot built with Power Virtual Agents.
- An engagement hub that is being used by live agents, such as Omnichannel for Customer Service, and you need to configure the connection, as described in Configure hand-off to Omnichannel for Customer Service.
Triggering hand-off to a live agent
Customers engaging with the bot can ask for a live agent at any point in the conversation. This escalation can happen in two ways, with an implicit trigger or an explicit trigger.
Upon triggering the hand-off topic, Power Virtual Agents starts the hand-off to the configured engagement hub and sends over all conversation context to find the next best live agent to ramp them up so they can resume the conversation.
In some instances, the bot may be unable to determine the intent of a customer's conversation. For example, the customer may be asking a specific question for which there's no topic, or no matching option within a topic.
In other instances, your customers may ask to be handed off to a live agent immediately. For example, customers may type "talk to agent" mid-way into a conversation.
When the bot detects an escalation in this manner, it will automatically redirect the user to the Escalate system topic. This type of trigger is known as implicit triggering.
When creating topics for your bot, you may determine that some topics require interaction with a human. This type of trigger is known as explicit triggering.
In these instances, you must add a Transfer to agent node into the topic.
This node lets you add a Private message to agent, which is sent to the connected engagement hub to help the live agent understand the history and context of the conversation.
Conversations that reach this node will be marked as Escalated sessions in reporting analytics.
Add a "Transfer to agent" node into a topic:
Go to the Topics page for the bot you want to edit.
Open the authoring canvas for the topic you want to add the Transfer to agent node to.
Select Add node (+) to add a message node. Enter what the bot should say to indicate that transferal to a live agent is about to occur.
Underneath the message node, select Add node (+), go to End the conversation and then select Transfer to agent
Enter an optional private message to the live agent in the Transfer to agent node. This optional message can be useful if you have multiple topics with Transfer to agent nodes as the information is stored in the
The topic will start the transfer to a live agent when this node is reached. You can test the hand-off by triggering the topic in the test canvas.
Once you add a Transfer to agent node into a conversation, each time you trigger hand-off your users will see a "No renderer for this activity" message on the demo website. This message suggests the need to customize your chat canvas to implement custom client-side code that brings in a human agent from your engagement hub into the conversation.
Contextual variables available upon hand-off
Beyond providing an automated way for a conversation to be ported into an engagement hub, it's important to ensure that the best agent for a specific problem is engaged. To help route conversations to the most appropriate live agent there are context variables that are also passed to the engagement hub.
You can use these variables to automatically determine where the conversation should be routed. For example, you may have added Transfer to agent nodes to several different topics, and you want to route conversations related to certain topics to specific agents.
The following table lists the context variables available by default.
||Route escalations to a live agent.||
||Route escalations to a live agent and help them ramp-up. Includes the last topic that was triggered by an utterance from the user.||
||Ramp-up a live agent. Only includes topics triggered by end user using a trigger phrase. Doesn't include topics that were redirected to.||
||Route escalation to a live agent and help them ramp-up.||
||Ramp-up a live agent.||
||Uniquely identify a bot conversation.||
||Ramp-up a live agent.||
||Identify the bot that is handing off a conversation.||
||Route escalation to a live agent.||
|All user-defined topic variables||Ramp-up a live agent.||
A customer may go through several topics prior to escalating. Power Virtual Agents gathers all contextual variables across topics and merges them before sending to the engagement hub.
If there are topics with similarly named context variables, Power Virtual Agents promotes the most recently defined topic variable.