Live Agent Transfer

The transfer of conversation between Copilot and an End User to a Live Agent (Analyst) is known as a "Live Agent Transfer" (LAT).

LAT basically provides an option to the end user to invoke Live Agents during the conversation with Copilot. This enables the end user to escalate or provide immediate attention to a logged service request (SR) or an incident (INC) by the assigned analyst.

LAT acts as a bridge that fills the gap between the end user and the analyst. It builds an immediate connectivity between the two.

When a Copilot cannot accurately respond to an end user's queries or when the end user requests that the conversation be transferred to an analyst, the Copilot may transfer the conversation to the analyst using the LAT.

Live Agent Transfer increases customer satisfaction since a live analyst can promptly respond to end user queries by providing the most relevant resolutions. Also, the end user can view the full history of their previous conversations.

The end user can continue to interact with Copilot, even after the Live Agent ends the conversation. If there are no further queries, the end user can end the conversation with the Copilot.

Skill-based routing in Live Agent Transfer

When an end user is conversing with the Copilot and requests for Live Agent Transfer, by default, based on the configuration of end user location and the Workgroup location, the routing of LAT happens. However, the Copilot can also connect to a Live Agent based on the skills.

Whenever an end user provides an input to the Copilot, the Copilot identifies the entity in the input. Accordingly, the Copilot routes the conversation to a Live Agent whose skills match the identified intent.

With skill-based routing Live Agents will be able to provide tailored assistance, understand the customer's requirements better and offer specialized solutions or recommendations.

Note:

Live Agent can access the Live Chat iconand accept the chat request from end user through Apex.

Suppose the end user provides an input ‘human agent’, 'live agent', or 'agent'. Copilot identifies the entity from the input provided by the end user. The end user wants to talk to a Live Agent.

Below is a sample screenshot:

Figure: Copilot confirms LAT

Copilot transfers the conversation to the Workgroup whose skill matches the identified entity and to an Analyst (Live Agent) who belongs to that Workgroup.

Below is a sample screenshot:

Figure: Copilot transfers the conversation to Live Agent

Transferring Copilot-End user conversation to Live Agent

Copilot works in two different ways to transfer the conversation to the Live Agent. It either directly provides the option of LAT to the end user or works as per the request of the end user.

Transfer of the conversation between Copilot and End User to the Live Agent happens in either of the following ways:

1. Copilot Transfers the Conversation to Live Agent
   
   When an end user begins a conversation with Copilot, the Copilot tries to provide a response/solution to the end user’s queries. If the end user is not satisfied with the response/solution provided by the Copilot for three consecutive times, then the Copilot asks whether the end user wants to transfer the conversation to an analyst. If the end user says ‘Yes’, then the conversation on the Copilot interface is retained and the conversation is transferred to an available analyst.
   
   
2. End User Wants to Talk to the Live Agent
   
   If an end user wants to talk directly to the Live Agent even before conversing with the Copilot, the Copilot connects the end user with the Live Agent. As in the following use case, the end user provides an input - 'live agent' - in Copilot without having any conversation with Copilot. So, Copilot then transfers the conversation to Live Agent.Figure: End user confirms Live Agent Transfer

LAT flow for Live Agent

Once the Copilot transfers the conversation to Live Agent, the Live Agent will need to perform certain steps to start the conversation with the end user.

To start conversation with an end user, perform the following steps:

1. The Live Agent logs into the Apex with Analyst access in order to participate in the conversations transferred from the Copilot to the Live Agent.
2. Use the Live Chat iconon the Analyst Dashboard, the Live Agent (Analyst) can 'Accept' or 'Ignore' the conversation transfer request from the end user. Refer the below screenshot:
   Figure: Live Agent icon
3. Once the Live Agent clicks on , a new tab will open in the browser (refer the below screenshot), where the Live Agent can accept or reject the chat transfer request, view history log of conversations, and user details.
   Figure: Live Agent Hub screen
4. Live Agent will see all the new conversation transfer requests under the New Requests panel. Once the Live Agent clicks on the end user name under the New Requests panel, the Live Agent Hub window will display all the details related to that specific end user. The right side of Live Agent Hub window will show the options to Ignore or Accept the conversation transfer request. (Refer the below screenshot)
   Figure: Ignore or Accept chat transfer request
5. The middle panel of the Live Agent Hub screen displays the initial conversation that happened between Copilot and end user. When the Live Agent accepts the Live Agent Transfer request, the following toast message is displayed stating that the Live Agent is connected with the End User.
   Figure: Chat accepted
   The conversation between the Live Agent and end user is displayed in the middle panel.
6. After the Live Agent accepts the conversation transfer request and starts the chat, the Live Agent Hub will have an End option (refer the below screenshot). The Live Agent can click End to end the chat.
   Figure: Live Agent Hub - end chat option

Live Agent Hub

The Live Agent Hub acts as a unified platform for Live Agents to view and manage conversations and conversation transfer requests. This hub allows Live Agents to access user details, historical conversation logs, and relevant data in real-time, facilitating a more informed response.

Live Agents have access to the following details in the Live Agent Hub window:

1. New Requests: New conversation requests from the end users appear on the New Requests panel (refer to the screenshot below). Live Agents can navigate to the New Requests panel to accept or reject the conversation transfer request.
2. Active Requests: All the active conversations which the Live Agent has accepted and having conversation appear on the Active Requests panel. The active conversations are displayed end user-wise along with the end user name. Click on the end user name to chat with a particular end user.
3. History: The History panel displays the last five conversation history of a Live Agent with the name of the Live Agent, date and time of the conversation. Click on a particular history detail to view the entire conversation. Even if the end user ends the chat, the details of the conversation will appear on the History panel. (Refer the below screenshot)
   Figure: Agent History
4. User Details: The User Details panel on the right side of the Live Agent Hub screen displays information about the end user like the company name to which the end user belongs, location and email address.
5. User's Record History: The User's Record History panel provides a log of the recent five tickets for the selected end user. There are two tabs in the User's Record History panel segregating Open and Closed tickets. (Refer the below screenshot)
   Figure: User's record history
   Once the Live Agent clicks on any ticket, it will open a new window where the details of the ticket can be found. The Incidents and Service Requests are displayed according to the Live Agent’s tenant access. For instance, if a Live Agent has IT tenant access but an Incident is generated in the HR tenant, that incident would not be visible to the Live Agent.
6. Action Panel: Live Agents can create an Incident or Service Request themselves through the Action Panel. (Refer the below screenshot). There are two tabs in the Action Panel – Create Incident and Create SR. Click on the respective tabs to log an Incident or SR.
   Figure: Action Panel
7. Eye Icon: The Eye icon near New Requests will enable Live Agents to quickly identify and understand which chats are being viewed by other Live Agents. (Refer the below screenshot)
   Figure: Eye icon in Live Agent Hub
   
   The Eye icon is visible to all other Live Agents (except the one who is viewing the chat) if any one Live Agent opens the chat request and is viewing it.
   
   If there is a new chat request from any end user, then the end user avatar under the New Requests section will have an orange border. (Refer the following screenshot)
   Figure: New chat request - orange border in End User avatar
   
   If any Live Agent has opened and viewed a chat request, then the respective border of the End User avatar under New Requests will change from orange to grey. The End User avatar of the chat requests which are not viewed by any of the Live Agents will have an orange border. (Refer the following sample screenshot)
   Figure: Grey border on End User avatar
   
   When the Live Agents hover-over the icon, then the tooltip will display the following information:
   1. If only one Live Agent is viewing the chat, then the tooltip will display ’Currently viewed by [Live Agent name]’ to other Live Agents. (Refer the sample screenshot)
      Figure: Eye Icon tooltip
   2. If there are two new chat requests and one is being viewed by a Live Agent, then all other Live Agents will see the Eye icon against the respective End User avatar. (Refer the following screenshot). Other chat requests which are not being viewed by any Live Agent will not have the Eye icon.
      Figure: Eye icon for chats viewed by Agents
   3. If more than one Live Agents are viewing the chat, then the tooltip will display 'Currently viewed by [no. of Agents] other Agents' to all other Agents who are not viewing the chat. (Refer the following screenshot)
      Figure: Eye icon tooltip displaying the number of Agents viewing the chat
      

Note:

If Live Agent logs out of the Apex, then the Live Agent is automatically logged out of the Live Agent hub window and the below message is displayed.
Figure: Logged out message

How to connect with Live Agent?

Copilot helps end user to connect with a Live Agent through Live Agent Transfer. The end users can check log an INC or SR, check the status of the INS or SR, or report any other concern with the Live Agent.

To connect with the Live Agent, the end user needs to take the following steps:

1. When the end user starts a conversation with Copilot, the Copilot displays a welcome message and provides options to the end user to log an Incident or Service Request (refer the below sample screenshot). These options are configurable.Figure: Copilot displaying the options to the end user
2. Copilot searches and provides response of the found catalog. Refer the below screenshot:
   Figure: Catalog display
3. The end user, however, wants to talk to the Live Agent. The end user provides an input to talk to the Live Agent.Figure: End user requests Live Agent Transfer
4. The Copilot confirms whether the end user wants to connect with the Live Agent. If the end user says ‘Yes’, the Copilot transfers the conversation to the Live Agent.Figure: Conversation transferred to Live Agent
5. Once the end user starts conversation with a Live Agent, an End Chat button appears, which the end user can click on to end the chat with the Live Agent. (Refer the below screenshot)Figure: End Chat
   
   After the end user clicks End Chat, the following message is displayed by Copilot:Figure: Reset conversation