Space banner

 

This version is currently available to SaaS customers only. It will be available to on-premises customers soon.

Sending chat transcripts to users via email

As an administrator, you can enable end users to receive the transcripts of their chat conversations at their email addresses that are registered with the organization. End users can get the transcript by entering a phrase in the chat. A transcript contains the complete conversation between a user and chatbot, along with the date and time of the conversation.

To enable users to receive a chat transcript, you must create the following items in the IBM Watson Assistant:

  • Intent—defines the action to perform such as create a ticket, find an issue.
  • Entity—is a class of object on which the action should be performed such as tickets and issues.
  • Dialog node—uses the intents and entities to interact with an application user.

To create an intent and entity to email a chat transcript

  1. Log in to IBM Watson Assistant and navigate to the Intents tab.
  2. Create the emailtranscript intent and provide examples such as I need the transcript of this chat or Send chat transcript.
  3. Click the Back option.
  4. Select Entities > MyEntities.
  5. Create the emailTranscript entity, and add now and later as the entity values, as shown in the image below:

To create a dialog node to email a chat transcript

  1. Log in to the IBM Watson Assistant and navigate to the Dialogs tab.
  2. Create the BMC Email Chat Transcript dialog.
  3. In the conditions, add the emailTranscript intent name.
  4. Under Then set context, for the $ sys_options variable, enter the following value:
    [{"label":"Send now","value":"now"},{"label":"Send later","later":"no"}]
  5. Under Assistant responds, under Text, enter the following text that the user sees as a response:
    I can send the transcript of the chat now, which will end the chat, or I can send it after completing the chat.
    The following image shows the configuration for the BMC Email Chat Transcript node:
  6. Create the child dialog nodes and configure them as specified in the following table:

    Child dialog nodeAction
    Send now
    1. In If assistant recognizes, enter emailTranscript:now.
    2. In the JSON editor, enter the following content:

      {
        "output": {
      
          "text": {
      
            "values": [
      
              "Chat transcript sent to
      $primaryEmailAddress. Have a good day"
      
            ],
      
            "selection_policy":
      "sequential"
      
          },
      
          "action": {
      
            "inputMap": {
      
              "chatId":
      "$chatId",
      
              "enableChatTranscript": true
      
            },
      
            "actionTypeName":
      "shouldSendChatTranscript",
      
            "waitForActionExecution": true
      
          }
      
        },
      
        "context": {
      
          "bmcSessionEmailChat": true
      
        }
      
      }

      The following image shows the configuration for the Send now node:

    3. For the Send now dialog, create the Close chat child dialog node, and specify the following details for the Close chat node:

      • If assistant recognizes—Enter True.
      • JSON editor—Enter the following content:

        {
        
          "output": {
        
            "text": {
        
              "values": [
        
                ""
        
              ],
        
              "selection_policy":
        "sequential"
        
            },
        
            "action": {
        
              "actionTypeName":
        "clearChatContext"
        
            }
        
          }
        
        }

        The following image shows the configuration for the Close chat node:

    Send Later
    1. In If assistant recognizes, enter emailTranscript:later.
    2. In the JSON editor, enter the following content:

      {
      
        "output": {
      
          "text": {
      
            "values": [
      
              "Okay, I'll send it later. You can
      continue with the chat now."
      
            ],
      
            "selection_policy":
      "sequential"
      
          }
      
        },
      
        "context": {
      
          "bmcSessionEmailChat": true
      
        }
      
      }

      The following image shows the configuration for the Send later node:

    Start over
    1. In If assistant recognizes, enter emailTranscript:!@start-over.
    2. In the JSON editor, enter the following content:

      {
      
        "output": {
      
          "generic": [
      
            {
      
              "response_type":
      "text",
      
              "values": [],
      
              "selection_policy":
      "sequential"
      
            }
      
          ]
      
        },
      
        "context": {
      
          "sys_options": []
      
        }
      
      }
    3. In Then assistant should, select Jump to > BMC Email Chat Transcript (Evaluate responses), as shown in the following image:

    You have successfully created the dialog nodes to enable users to receive a chat transcript by entering the required phrase in the chat. 
    The following image shows how the bot prompts a user to specify when the user wants the chat transcript:

Related topics

Submitting service requests or searching knowledge articles

Using live chat to escalate issues to support agents

Was this page helpful? Yes No Submitting... Thank you

Comments