Lab 4 - Digital Deflection

Objectives

In this lab, you will learn how to integrate an Autonomous AI agent into a chat flow.

Know how to create an Autonomous AI Agent from scratch and map it to a knowledge base.
Create an agent goal, instructions, and welcome message.
Understand how to configure and use fulfillment.
See how to integrate the AI Agent into a call studio script.

Explanation of formatting:

Italics - This indicates a button, menu option, or object that should be selected.
"Quoted Text" - This indicates that the text should be typed exactly as indicated.

Instructor Demonstrated

Note, this lab is going to be instructor demonstrated. There are no tasks that you will do in this lab. The lab guide is provided for you to refer to later.

Section 1 - Add Autonomous AI Agent to Inbound chat flow

In this section we will integrate the Autonomous AI agent to a inbound chat flow. You will not perform any of the steps in this section, they are simply for your reference.

Task 1. Import Inbound Flow

Step 1. Download a customized Inbound flow by selecting the link. Unzip this to a location on your computer.

Step 2. Login to Webex Connect by navigating to the Contact Center -> Digital menu options in Webex Control Hub.

Step 3. Ensure that you have created a Live Chat Asset and Service to use. You can find instructions here.

Step 4. In the service you created, select Flows, then select the Create Flow button.

Provide a descriptive name for the Flow Name.
From the Method drop-down, select "Upload a flow"
Either drag and drop the extracted flow into the Attachment box or, select Choose File and navigate to where you extracted the zip file and choose the .workflow file.
Select Create to upload this flow and open it.

Step 5. Modify the flow.

a. Accept the defaults in the initial Configure APP Event.

b. Locate the AI Agent Node and open it.

c. In the AI Agent Node, select Autonomous from the Agent Type drop-down box. Next, select the AI Agent you created in the Agent drop-down.

Note the Message is a variable. Each time that the loop occurs, this message will be updated with the customer's response back to the agent. This variable should be set to a message which will start the agent such as "Hello".

d. Review the remaining nodes in the workflow. These are as follows:

Evaluate - This simply sets the timestamp to the current time.
Sent agent response text - This will display the response from the agent to the customer.
Append Conversation - This will write the agent response to history.
Receive - This receives the response from the customer back to the agent.
Evaluate - This normalizes the response from the customer and ensures there are no content issues. This will then set the input message variable to the sanitized response.
Append Conversation - This writes the customer's response to history.

e. At any point if the customer asks to speak to an agent, the AI Agent Node will take the onAgentHandover exit path and will be sent to CCE for agent handling.

Task 2. Save and make Live

Step 1. Save the Flow.

Save your changes by selecting Save at the top of the screen. If any messages come up that nodes are not configured, select each message, then check to ensure that the node is correctly configured and saved. Some nodes simply need to be opened and saved after a workflow has been imported.

Step 2. Make Flow live and map to an application.

Select the Make Live button at the top of the work flow. In the Make Live Configuration dialog box, select the application you wish to map this flow to. It will take several seconds to a minute for the flow to go live.

Task 3. Install to Chat

The steps to customize the chat interface and install in a web site will not be covered in detail in this section.

You can find details at this link

Section 2 - Scripted AI Agent Steps

Note: This is not going to be shown in the class but is here so you can see what changes need to be made to the AI agent to add in chat support.

Task 1. Create Intents

In this task, we will create the following 2 Intents

Chat Start Order Tracking
Chat Order Results

Step 1. Create Intent - "Chat Start Order Tracking"

a. In the AI Agent configuration, navigate to Script -> Intents -> click on the Create intent button.

b. Enter the Intent name as "Chat Start Order Tracking".

c. Under Intent and Utterances section, type the phrase "Start_Chat" and click the Add button.

Note:

For the Voice channel, the welcome_event was sent from the Call Studio script as the trigger for the AI Agents' welcome message.
Similarly, for the Web (Chat) channel, the phrase "Start_Chat" will be used as the trigger for the AI Agents' welcome message. This will be configured in the Inbound Flow in a later step.

d. In the Response section, from the drop down, select Welcome message.

e. Lastly, click Create to complete the "Chat Start Order Tracking" intent configuration.

Step 2. Create Intent - "Chat Order Results"

a. In the AI Agent configuration, navigate to Script -> Intents -> click on the Create intent button.

b. Enter the Intent name as "Chat Order Results".

c. Under Intent and Utterances section, type the phrase "order_results" and click the Add button.

Note:

In the Inbound Flow, once the order_number and user_pin slots are collected as part of the "Start Order Tracking" Intent, the Inbound Flow sends the order_results message to trigger the track_order_results response.

d. In the Response section, from the drop down, select track_order_results.

e. Lastly, click Create to complete the "Chat Order Results" intent configuration.

Task 2. Update Responses for the Default channel (web)

In this task, we will be updating the following responses - Default channel (web) - Chat

Welcome message
order_number
pin_missing
track_order_wait
track_order_results
track_another_order
Goodbye

Step 1. Update the Welcome message response

a. In the AI Agent configuration, navigate to Script -> Responses -> click on the Welcome message response.

b. On the left side of the response window, click on the previously created conditional response "check_if_name_is_passed" -> Select the Rules tab.

c. Click the +OR Condition button and enter as below:

Left variable: consumerData.extra_params.name
Operator: Exists

c. Next, select the Actions tab on the left and click the Default (web) tab (Note: since this is for the chat channel, ensure you are NOT selecting Voice option)

d. Add the text below as Variant 1 in the Text section. Feel free to add more of these variants if you like.

Note: consumerData. is how we can reference data that is passed from Chat Flow to the bot.

Hello  ${consumerData.extra_params.name} , I am a Tracker bot. I can help you track your packages. How can I help you today? To get started, please ask me to track an order number.

Step 2. Update the order_number response

a. In the AI Agent configuration, navigate to Script -> Responses -> click on the order_number response.

b. Next, click on the Default (web) button and enter the below text as Variant 1.

Please provide your six digit order number.

Step 3. Update the pin_missing response

a. In the AI Agent configuration, navigate to Script -> Responses -> click on the pin_missing response.

b. Next, click on the Default (web) button and enter the below text as Variant 1.

Please provide your four digit PIN.

Step 4. Update the track_order_wait response

a. In the AI Agent configuration, navigate to Script -> Responses -> click on the track_order_wait response.

b. Next, click on the Default (web) button and enter the below text as Variant 1.

Getting your package details now...

Step 5. Update the track_order_results response

a. In the AI Agent configuration, navigate to Script -> Responses -> click on the track_order_results response.

b. On the left side of the response window, click on the previously created conditional response "order_status_check" -> Select the Rules tab

c. Click the +OR Condition button and enter as below:

Left variable: consumerData.extra_params.order_status
Operator: Exists

c. Next, select the Actions tab on the left and click the Default (web) tab (Note: since this is for the chat channel, ensure you are NOT selecting Voice option)

d. Add the text below as Variant 1 in the Text section. Feel free to add more of these variants if you like.

Note: consumerData. is how we can reference data that is passed from Chat Flow to the bot.

We have found your order. ${consumerData.extra_params.order_status}  Do you need anything else?

Step 6. Update the track_another_order response

a. In the AI Agent configuration, navigate to Script -> Responses -> click on the track_another_order response.

b. Next, click on the Default (web) button and enter the below text as Variant 1.

Great! To get started, please ask me to track an order number.

Step 7. Update the Goodbye response

a. In the AI Agent configuration, navigate to Script -> Responses -> click on the Goodbye response.

b. Next, click on the Default (web) button and enter the below text as Variant 1.

Thank you for your response! I will now connect you to a live agent for further assistance! Have a wonderful day!

Task 3. Import and Configure the Native AI Inbound Chat Flow

Step 1. Navigate to this link. Unzip this to a location on your computer.

Step 2. Log into Webex Control Hub admin.webex.com and navigate to Contact Center -> Digital -> Webex Connect. This opens up the Webex Connect portal.

Step 3. Navigate to Services tab and select the Service created for you seat number.

Step 4. Click the Create Flow button -> enter a Flow Name -> select Upload a Flow -> Choose the downloaded file -> Click Create

Step 5. Update the Bot Start and the Bot OrderResults nodes

Under the Agent field, select the AI Agent Bot you created

Step 6. Review the nodes to see where the messages are sent to the AI Agent.

Step 7. Click on the Settings tab and update the following variables:

CloudConnectFQDN
botInputMessage

Step 8. Save and Make Live

Task 4. Install to Chat

The steps to customize the chat interface and install in a web site will not be covered in detail in this section.

You can find details at this link

This now completes lab 4