Overview

The following documentation outlines the configuration for the connector and how to implement functions specifically for IBM Watson.

Bot Configuration

See the Getting Started guide first to complete pre-requisite steps.

Please note that Watson does not support processing newline, tab and carriage-return characters. These symbols will be removed from any query that is sent to Watson via the provided connector.

With watson there are two ways of authentication that currently our system support, these are UserPass and IAM (token based) authentication. You can choose one of them for your bot configuration.

UserPass authentication

You will be presented with following screen to complete the Vendor Settings in order to add bot connector using UserPass authentication.

Figure 1.1 Showing the configuration that needed to be filled using UserPass authentication

Following information needs to be completed for LivePerson:

IAM authentication

You will be presented with following screen to complete the Vendor Settings in order to add bot connector using IAM authentication.

Figure 1.2 Showing the configuration that needed to be filled using IAM authentication authentication

Following information needs to be completed for LivePerson:

Test Connection

You have to agree to Data Disclaimer from now onward in order to use the services of bot connector. For that you can click on the checkbox "I agree to the Data Disclaimer

For validation of the credentials provided, you can now perform a test connection request to see if everything that you have provided is working and reachable. You can click on the button "Test Connection" to see if connection succeed or fail. For UserPass authentication see in Figure 1.3 and 1.4. For IAM authentication see in Figure 1.5 and 1.6.

Figure 1.3 Showing the success case of the valid credentials for UserPass authentication

Figure 1.4 Showing the fail case of the invalid credentials for UserPass authentication

Figure 1.5 Showing the success case of the valid credentials for IAM authentication

Figure 1.6 Showing the fail case of the invalid credentials for IAM authentication

Once you are done with providing configuration you can save it by pressing on "Done". Congratulations! You have completed the configuration of the Watson Assistant bot.

Following guide is going to present customization for the Watson Assistant on how to implement functions specifically for IBM Watson. It is intended for users who are familiar with IBM Watson cloud dashboard. Continue if you are familiar and have access to IBM Watson cloud dashboard.

Sending Native Content

Watson Assistant allows the user to define native response types to the dialog nodes. The supported Watson Assistant native types include Image, List, Pause, and Text. Users can define single or multiple native content per dialog. The native content types can be defined with Watson wizard or using the JSON editor (Figure 2.1 shows how to access both ways in IBM Watson website).

Figure 2.1 IBM Watson Native Rich Content Wizard and JSON Editor

Please note that Watson assistant API version of 2018-09-20 is used to support the native content response in Third-Party Bots.

If you use JSON Editor then the usual body of the native content is as follows:

{
  "output": {
    "generic": [
      // Here comes array of objects of different Watson native contents that you can define.
    ]
  }
}

Image

User can define Image type using the IBM watson assistant dashboard. To do this, dialog node will need to selected that will hold image response. Click on the "Add response type" and select Image from the select box as shown in Figure 2.1.

Figure 2.1 Response type of Image is highlighted

Once image is selected you will be asked to fill the information. "Image Source" url must be provided. You can also describe the image title and description (example filled form is shown in the Figure 2.2).

If Images are sent in Rich content, then their URLs must be added to a whitelist via internal LivePerson configuration (Houston: messaging.rich.content.valid.urls). Please note that you must add all possible domains to this list manually as wildcards are not supported. Moreover, All domains must be HTTPS secure.

Figure 2.2 Image fields filled example

If you are using JSON editor you can add a Image type by posting following JSON. Please make sure to change "source", "title" and "description" property with your data.

{
  "output": {
    "generic": [
      {
        "source": "https://images.pexels.com/photos/699122/pexels-photo-699122.jpeg",
        "title": "iPhone 8 Concept",
        "description": "iPhone 8 concept image showing initial details about phone",
        "response_type": "image"
      }
    ]
  }
}

List

User can define List type using the IBM watson assistant dashboard. To do this, dialog node will need to selected that will hold list response. Click on the "Add response type" and select Option from the select box as shown in Figure 2.3.

Figure 2.3 Response type of List is highlighted

Once the "Option" is selected the form need to be filled will be shown. You must provide "Title" and also "Description". Furthermore, different choices of options can be added via clicking "Add option" button. Once the button is clicked you will be asked to put a label of option and value. Make sure you fill both of them (example filled form shown in Figure 2.4).

Figure 2.4 List fields filled example

If you are using JSON Editor then you have following structure of List. Note that "options" property is array of objects which holds the items for choosing are presented to user.

{
  "output": {
    "generic": [
      {
        "title": "",
        "description": "",
        "options": [
          // Here comes the list of options you want to present to user
        ],
        "response_type": "option"
      }
    ]
  }
}

An example list filled with two options can be seen below. Please note that within options object, "text" (value->input->text) is the value that you set for an option.

{
  "output": {
    "generic": [
      {
        "title": "Choose your Phone",
        "description": "Select the phone you like",
        "options": [
          {
            "label": "iOS",
            "value": {
              "input": {
                "text": "1"
              }
            }
          },
          {
            "label": "Android",
            "value": {
              "input": {
                "text": "2"
              }
            }
          }
        ],
        "response_type": "option"
      }
    ]
  }
}

Pause/Delay

Users can define Pause type if they want to send some delay in responding. For adding this content type, the dialog node will need to select that will hold pause response. Click on the "Add response type" and select Pause option as shown in Figure 2.5

Figure 2.5 Response type of Pause is highlighted

Once the "Pause" is selected the form will ask you to provide the duration (unit is in milliseconds). This allows the conversation to be paused for the amount of time defined in "Duration" field. Moreover, If you want to show user a indication of typing you can select choose that with Typing Indicator radio box. (example filled form is shown in Figure 2.6). This will show a indication like "Agent is typing…" for the amount of time of delay that is set in "Duration".

Note: using the delay as a single/sole response from the bot to the consumer, is effectively a ‘no response’ action. Using this allows the bot to receive a consumer message without responding to the consumer.

Figure 2.6 Pause fields filled example

If you are using JSON Editor you can use the following JSON structure to define a Pause content type. This example will pause for 5 milliseconds with typing indication on.

{
  "output": {
    "generic": [
      {
        "time": 5,
        "typing": true,
        "response_type": "pause"
      }
    ]
  }
}

Text

Users can define a Text type to send some textual response. For adding this type dialog node will need to select that will hold text response. Click on the "Add response type" and select "Text" option as shown in Figure 2.7

Figure 2.7 Response type of Text is highlighted

Once the "Text" is selected the form will allow you to add the response texts. You can add multiple responses variation (example filled form is shown in Figure 2.8).

Figure 2.8 Text fields filled example

If you are using JSON Editor you can use following JSON structure to create text responses. The example below shows two text responses defined that will come sequentially.

{
  "output": {
    "generic": [
      {
        "response_type": "text",
        "values": [
          {
            "text": "Hi Good Morning!"
          },
          {
            "text": "Hi Good Evening!"
          }
        ],
        "selection_policy": "sequential"
      }
    ]
  }
}

Defining multiple responses with Watson Native content

Users can define a response with various content types. The following example shows a similar case using JSON Editor. The response will First send the text. afterwards, it will make a pause for 5 seconds and then finally sends an image.

{
  "output": {
    "generic": [
      {
        "values": [
          {
            "text": "Hi Good Morning!"
          },
          {
            "text": "Hi Good Evening!"
          }
        ],
        "response_type": "text",
        "selection_policy": "sequential"
      },
      {
        "time": 5000,
        "typing": true,
        "response_type": "pause"
      },
      {
        "title": "iPhone 8",
        "source": "https://cdn.bgr.com/2016/08/iphone-8-concept.jpg",
        "description": "iPhone 8 concept",
        "response_type": "image"
      }
    ]
  }
}

Sending Rich Content (Structured Content)

Please note that Watson assistant API version of 2018-09-20 is used to support the rich content response in Third-Party Bots.

The core LiveEngage platform supports the use of rich/structured content. For more information on the format and functionality available, please refer to the documentation found here. As a result, the Bot Connector also supports this.

To send structured content via Watson Assistant you will need send custom JSON. To do this, you will need to select the dialog node that will hold the structured content (Figure 3.1).

Figure 3.1 Watson Dialog Node

From there, under the section Then respond with: Click the three vertical dots and select Open JSON Editor (Figure 3.2)

Figure 3.2 Watson Assistant Dialog JSON Editor

In the JSON Editor you will need to add your custom JSON response (Figure 3.3).

Figure 3.3 Watson Assistant JSON Editor

There is a strict JSON structure for the response that must be used. The JSON structure can be found below in Figure 3.4. An example with a sample JSON that uses a standard Structured Content card with a button option in can be seen in Figure 3.5.

{
  "output": {
    "text": {
      "values": [
        {
          "metadata": [
            {
              "id": "1234",
              "type": "ExternalId"
            }
          ],
          "structuredContent": {}
        }
      ],
      "selection_policy": "sequential"
    }
  }
}

Figure 3.4 Structured Content Watson JSON Structure (JSON Editor should contain this object structure for Rich Content)

{
  "output": {
    "text": {
      "values": [
        {
          "metadata": [
            {
              "id": "1234",
              "type": "ExternalId"
            }
          ],
          "structuredContent": {
            "type": "vertical",
            "elements": [
              {
                "type": "button",
                "click": {
                  "actions": [
                    {
                      "text": "Recommend me a movie, please",
                      "type": "publishText"
                    }
                  ]
                },
                "title": "Recommend a movie"
              }
            ]
          }
        }
      ],
      "selection_policy": "sequential"
    }
  }
}

Figure 3.5 Structured Content Watson JSON Example (JSON Editor should contain this object structure for Rich Content)

For new IAM workspaces that have a new Watson response, Then respond with "text" should be used:

Put the structured content objects that is shown in Figure 3.6 with the metadata in the text field. Figure 3.7 shows the final picture of how it should look like.

{
  "metadata": [{ "id": "1234", "type": "ExternalId" }],
  "structuredContent": {
    "type": "vertical",
    "elements": [
      {
        "type": "button",
        "click": {
          "actions": [
            { "text": "Recommend me a movie, please", "type": "publishText" }
          ]
        },
        "title": "Recommend a movie"
      }
    ]
  }
}

Figure 3.6 Structured Content Watson JSON Example (IAM)

Sending Quick Replies (Structured Content)

Please note Quick Replies are only supported in Messaging Conversations.

Quick Replies are a special kind of the Structured Content. The content should be added to the quickReplies property of the structuredContent object and there also a message be included in the structuredContent object. This message will be sent to the customer along with the Quick Replies. For detailed information on Quick Replies check out the documentation for the specific channel (Mobile SDK and Web, Facebook Messenger, Google RCS Business Messaging).

{
  "structuredContent": {
    "quickReplies": {
      "type": "quickReplies",
      "itemsPerRow": 8,
      "replies": [
        {
          "type": "button",
          "tooltip": "yes i do",
          "title": "yes",
          "click": {
            "actions": [
              {
                "type": "publishText",
                "text": "yep"
              }
            ],
            "metadata": [
              {
                "type": "ExternalId",
                "id": "Yes-1234"
              }
            ]
          }
        },
        {
          "type": "button",
          "tooltip": "No!",
          "title": "No!",
          "click": {
            "actions": [
              {
                "type": "publishText",
                "text": "No!"
              }
            ],
            "metadata": [
              {
                "type": "ExternalId",
                "id": "No-4321"
              }
            ]
          }
        }
      ]
    },
    "message": "Message to send before sending QuickReplies content"
  },
  "metadata": [
    {
      "id": "1234",
      "type": "ExternalId"
    }
  ]
}

Figure 3.7 Quick Replies StructuredContent example.

Change Time To Response of Conversation

Change the TTR of a conversation based on the action response of Watson. There have 4 different types. "URGENT", "NORMAL", "PRIORITIZED", "CUSTOM". Only the "CUSTOM" can set a value. The unit of the value is second. And the value of the others are defined in the Agent Workspace.

{
  "output": {
    "text": {
      "values": ["Sure thing! Change the TTR to 50 minutes."],
      "selection_policy": "sequential"
    }
  },
  "actions": [
    {
      "name": "CHANGE_TTR",
      "type": "CLIENT",
      "parameters": {
        "ttrType": "CUSTOM",
        "value": 3000
      },
      "result_variable": "none"
    }
  ]
}

Figure 4.1 Watson JSON response for changing TTR

Transfer/Escalations

Transfers and escalations are straightforward in both chat and messaging. At the beginning of a chat session or when a messaging bot logs in, all the list of enabled skills on the account are retrieved, keyed by name and stored. When a transfer is requested by the bot, the skill name is matched to one already on the account and the id is retrieved and escalated to. In regards to Watson Assistant, this should be configured in the following way:

In the Then respond with: JSON editor block, we see the following:

{
  "output": {
    "text": {
      "values": ["Escalating to a human"]
    }
  },
  "actions": [
    {
      "name": "TRANSFER",
      "type": "client",
      "parameters": {
        "skill": "human_skill"
      },
      "result_variable": "none"
    }
  ]
}

Figure 5.1 Watson JSON response for escalation

Above is the actions array. Here, we have a escalation skill name in the skill parameter. This is the name of our skill for escalation. This will be sent in the BOSO object to the chat/messaging connector, which will grab the skillId from an array based on the name, and escalate.

Close Chat/Conversation

To close a chat or messaging conversation, we utilize the action object as we did for a transfer (see Figure 5.1). In Figure 6.1 below, the Watson Assistant JSON response should be mirrored as follows:

{
  "output": {
    "text": {
      "values": ["Thanks for chatting with us today!"],
      "selection_policy": "sequential"
    }
  },
  "actions": [
    {
      "name": "CLOSE_CONVERSATION",
      "type": "client",
      "result_variable": "none"
    }
  ]
}

Figure 6.1 Watson Assistant JSON response for closing chat/conversation

Engagement attributes as context

Third-Party bots allows the collection of engagement attributes (more information can be found here) if Engagement Attributes option is checked in the Conversation Type step as shown in Figure 7.1.

Figure 7.1 Conversation Type step in creation/modification of bot configuration.

These attributes are only collected at the start of a conversation. Third-Party bots leverage the LivePerson Visit Information API to collect the engagement attributes, Further information Visit Information API can be found here. Moreover, Engagement attributes are not updated throughout the life cycle of a conversation and only passed along with each message request. In Watson Assistant V1 these engagement attributes are added to the property lpSdes. For the preservation of these attributes within a conversation, context property is used (further information about context can be found here). An example of the request body can be seen below:

{
  "message": "Some Message",
  "context": {
    // ... contains some more information about conversation as well
    "lpEvent": {}, // Holds LP Events
    "lpSdes": {}
  }
}

Limitations

• Currently IBM Watson allows only 5 response types per node.

Sending Encoded Metadata

LiveEngage Messaging platform provides a new metadata input type (“encodedMetadata”) for passing a base64 encoded metadata on a conversation. The new metadata input type is in addition to the existing conversation metadata input field. Third-party Bot also supports this property and this section will cover the information needed for you to send encoded metadata within your conversations. Before sending encoded metadata you must ensure the following conditions in order to successfully send the data.

• Common.EncodedMetadata AC feature is ON
• Content is base64 encoded
• Metadata size is limited to 5k

Failing to comply with the above validation points will cause the message to be dropped. This feature is only available for the messaging conversations not for chat conversations

Encoded Metadata can be sent with simple Text, Rich Content (structured content) and Multiple responses. The encodedMetadata can be defined with context editor or using the JSON editor(Figure 8.1). In both ways, Third-Party Bot leverages the context variables to send the encoded metadata. The encoded metadata is passed throughout the conversation responses unless it is overwritten by upcoming encodedMetadata. Then the upcoming encodedMetadata will be passed along the responses.

Figure 8.1 Showing context editor with the encoded metadata.

Be careful with the camel-case characters encodedMetadata you must provide it exactly the same.

Sending Watson Native Content with Encoded Metadata

Sending encoded metadata with the Native Content (Text, Image and Options) is possible using Watson context editor or also through the JSON editor. An example response definition for both ways can be seen below:

Figure 8.2 Showing context editor with the encoded metadata.

Example response body for JSON editor:

{
  "output": {
    "generic": [
      {
        "values": [
          {
            "text": "This is text with encoded metadata"
          }
        ],
        "response_type": "text",
        "selection_policy": "sequential"
      }
    ]
  },
  "context": {
    "encodedMetadata": "ewoiaWQiOiAidGV4dCIKfQ=="
  }
}

Sending Rich Content (structured content) with Encoded Metadata

Sending encoded metadata with the Native Content is possible using Watson context editor or also through the JSON editor. An example response definition for both ways can be seen below:

Figure 8.3 Showing context editor with the encoded metadata.

{
  "output": {
    "text": {
      "values": [
        {
          "metadata": [
            {
              "id": "1234",
              "type": "ExternalId"
            }
          ],
          "structuredContent": {
            "type": "vertical",
            "elements": [
              {
                "type": "button",
                "click": {
                  "actions": [
                    {
                      "text": "Recommend me a movie, please",
                      "type": "publishText"
                    }
                  ]
                },
                "title": "Recommend a movie"
              }
            ]
          }
        }
      ],
      "selection_policy": "sequential"
    }
  },
  "context": {
    "encodedMetadata": "ewoiaWQiOiAic2MiCn0="
  }
}

If you have a different context for multiple dialogs in a conversation that are responding with structure content, then you can define the encodedMetadata for each of the dialog separately. In that case, you will not define encodedMetadata via context editor or inside context property if using JSON editor. Below example shows where encodedMetadata property should be placed in structured content. Please note you will have to use JSON editor for defining such case:

{
  "output": {
    "text": {
      "values": [
        {
          "encodedMetadata": "ewoiaWQiOiAic2MiCn0=",
          "metadata": [
            {
              "id": "1234",
              "type": "ExternalId"
            }
          ],
          "structuredContent": {
            "type": "vertical",
            "elements": [
              {
                "type": "button",
                "click": {
                  "actions": [
                    {
                      "text": "Recommend me a movie, please",
                      "type": "publishText"
                    }
                  ]
                },
                "title": "Recommend a movie"
              }
            ]
          }
        }
      ],
      "selection_policy": "sequential"
    }
  },
  "context": {}
}