1. REDIRECT 5.3:Tutorials-for-admins/AI/HowtoCreateWatsonAssistant

1. REDIRECT 5.3:Contact-center-administrator-guide/CallCenterConfiguration/IntegrationAccounts/BotChatSuggestionsEngine

1. REDIRECT 5.3:Contact-center-administrator-guide/CallCenterConfiguration/IntegrationAccounts/SpeechToText

1. REDIRECT 5.3:Contact-center-administrator-guide/CallCenterConfiguration/IntegrationAccounts/TextToSpeech

1. REDIRECT 5.3:Scenario-builder-reference-guide/ScenarioBlocks/SetVariable

1. REDIRECT 5.3:Scenario-builder-reference-guide/ScenarioBlocks/ChatBotSelectAccount

1. REDIRECT 5.3:Scenario-builder-reference-guide/ScenarioBlocks/PlayListen

1. REDIRECT 5.3:Scenario-builder-reference-guide/ScenarioBlocks/AskaBot

1. REDIRECT 5.3:Scenario-builder-reference-guide/ScenarioBlocks/AskaBot

1. REDIRECT 5.3:Scenario-builder-reference-guide/ScenarioBlocks/If

1. REDIRECT 5.3:Scenario-builder-reference-guide/ScenarioBlocks/Goto

1. REDIRECT 5.3:Scenario-builder-reference-guide/ScenarioBlocks/SetVariable

1. REDIRECT 5.3:Scenario-builder-reference-guide/ScenarioBlocks/WebScreenPop

1. REDIRECT 5.3:Scenario-builder-reference-guide/ScenarioBlocks/FindAgent

1. REDIRECT 5.3:Scenario-builder-reference-guide/ScenarioBlocks/Exit

1. REDIRECT 5.3:Scenario-builder-reference-guide/Exercises/ScenarioBuilderBasics

AI & Bots Overview

AI and Bot tutorials provide step-by-step instructions on how to create functional chatbots and integrate them into your contact center services.

Learn how to create chatbots, configure chatbot integration, and how to get your Bright Pattern scenarios to work with your integrated chatbots.

For more information about using the Contact Center Administrator application, see all Tutorials for Admins.

Bot Creation

• How to Create an Amazon Lex Bot
• How to Create a Watson Assistant

Integration Account Configuration

• How to Add a Bot or Chat Suggestions Engine
• Set up an AWS Lex Integration Account
• Set up a Watson Assistant Integration Account

Putting It All Together in Bright Pattern

• How to Integrate Bots with Chat

How to Create a Watson Assistant

Bright Pattern Contact Center integrates with providers like IBM Watson to enable chatbots to be used in your contact center services.

In this article, you will learn how to create a basic IBM Watson Assistant that can be used as a conversational bot with your configured chat service. Note that the instructions provided in this article apply to either IBM Watson Assistant or IBM Watson Assistant (Conversation).

Procedure

This procedure will walk you through the process of setting up your first Watson Assistant. For a deeper understanding of Watson Assistant and other IBM resources, refer to IBM’s Getting Started tutorial and IBM’s API Reference.

Step 1: Create an IBM account

1. If you haven’t already done so, create an IBM account. This process creates and activates an IBMid.
   
   
2. Sign up for IBM Cloud. IBM Cloud is where you will be developing and managing resources like Watson Assistant bots.

Step 2: Add Watson Assistant as a resource

1. Search IBM's catalog of resources for Watson Assistant.
   
   
2. Edit service name, region, and select a resource group.
   
   
3. Click Create Assistant to add Watson Assistant.
   
   
4. The Assistants page will open, showing the skills available for your account plan type.
   
   

Step 3: Add a dialog skill

Skills are the workspaces where you will be developing your bot (note that IBM skills used to be called workspaces). Skills are what provide Natural Language Understanding (i.e., sentiment analysis) for your Watson Assistant. Because you are making a conversational bot, you will be building a dialog skill for talking to customers during live chat.

1. Click Add dialog skill.
   
   
2. For the sake of this example procedure, select Use sample skill, and click Customer Care Sample Skill, which is already set up for you to use and edit. Click it again and the Assistant page will open. That is where you develop your skill.
   
   
   

   Use sample skill
   
   
   

Step 4: Create and/or add intents for your dialog skill

Because you selected the sample skill, the Assistant page will show some preconfigured intents. Intents are actions triggered by keywords entered by your customer.

In Watson syntax, intents always begin with the hashtag ("#") symbol, followed by word(s) in title case (i.e., where the first letter of a word is capitalized, unless it's a preposition like "to" or an article like "the"). Multiple words are separated by underscores ("_").

Like this: #Talk_to_Someone

Get familiar with intents by clicking on the first one in the list. In this example, we clicked on #Cancel to review its properties. You can leave all the preconfigured intents as-is or add new user examples.

Intent Properties

For every intent, you need to specify:

• Intent name - Descriptive name consisting of "#" and text (e.g., "#Cancel")
  
  
• Description - What this intent does (e.g., "Cancel current request")
  
  
• User examples - Keywords and phrases that trigger the intent (e.g., anything a customer might type during the chat); add as many examples and variations of spelling/phrases as you can, as the user examples are what make the bot compelling and helpful

Step 5: Create entities

Click on the Entities tab at the top of the Assistant page. As with intents, you will see some preconfigured entities. Entities are like word sets that more narrowly define the customer text that the bot recognizes.

In Watson syntax, entities always begin with the "@" symbol, followed by word(s) in lowercase. Multiple words are separated by underscores ("_").

Like this: @zip_code

Get familiar with entities by clicking on the first one in the list. In this example, we clicked on @holiday to review its properties. For this example, leave all the preconfigured entities as-is.

Entity Properties

For every entity, you need to specify:

• Entity name - Descriptive name consisting of the "@" symbol and text (e.g., "@holiday")
  
  
• Value name - A particular word that's included under the umbrella of the entity (e.g., "Hanukkah" and "Christmas" could be values for entity "@holiday")
  
  
• Synonyms - Related words that could be used in place of the value name (e.g., "turkey day" could be a synonym for "Thanksgiving"); if you don't know any synonyms, IBM can recommend some for you

Step 6: Design your dialog flow

Click on the Dialog tab at the top of the Skills page. The dialog for the sample skill you selected in Step 3 will be shown.

A dialog is like a scenario in that it defines what the bot does in response to a customer's text or actions. When you design your dialog flow, you are telling the Watson Assistant what to do when it recognizes defined intents and entities during an active chat. Branches of a dialog are called nodes, and nodes can be organized into folders.

Dialog Properties

In this example, we are going to leave everything in the dialog as-is. Click on the "Hours of Operation" node to open up properties for that node:

• Name - The name of the node (you can edit it here)
  
  
• If assistant recognizes - This is where you specify intents related to the node name (e.g., for this "Hours of Operation" node, you are telling the bot what to do when it recognizes the user examples defined in the "#Customer_Care_Store_Hours" intent)
  
  
• Then respond with - This is where you specify entities and how the bot should respond when it recognizes them
  
  
  • If assistant recognizes - Enter entities here (with "@" symbol)
    
    
  • Respond with - This is the bot's response to the customer; in Bright Pattern's Agent Desktop, anything entered in this field becomes the bot suggestion
    
    
• Wait for user input - Select this to provide a bot response only when the customer types something

Step 7: Try the Dialog

1. Click the Try it button at the top of the page. This button launches the Watson Assistant you just built within a chat window.
   
   
2. Pretend to be a customer and type some text into the text input field to see how the bot responds. The Try it out tool will show the intents and entities that the bot recognizes from the text you entered. You can click on the intents shown to select other ones or mark them as irrelevant.
   
   
   

   Sample dialog

Please read IBM's documentation to get a deeper understanding of Watson Assistant skills, intents, entities, and dialogs. Refer to IBM’s Getting Started tutorial and IBM’s API Reference.

Step 7: Add the Watson Assistant

Now that you're done creating a dialog skill, it's time to assign the skill to your Watson Assistant. This last step is important because later, you may have multiple Watson Assistants and skills, and IBM needs to know which ones are assigned to each other.

1. Click Add Assistant.
   
   
2. Select the Use sample skill tab, which is the skill you just set up in this procedure. If you do not see it listed, select the Add existing skill tab and find it there..
   
   
3. Click on the desired skill name. That's it.

Next Steps

You have now set up a very basic Watson Assistant that can be integrated with Bright Pattern Contact Center. You may now:

• Read IBM's documentation, explore the Watson Assistant, and edit it as desired
• Create a Watson bot/chat suggestions integration account
• Edit a chat scenario that uses your Watson Assistant
• Configure web chat to work with bots

How to Create an Amazon Lex Bot

Bright Pattern Contact Center integrates with Amazon Lex, a platform for building, testing, and deploying chatbots from the AWS Management Console. Lex provides both automatic speech recognition (ASR) and natural language understanding (NLU) technologies, enabling chatbots to recognize customers’ speech and text input, understand intent, and transcribe speech input. Integration with Lex lets your contact center access Lex through chat scenarios and provide bot assistance from within chat interactions.

In this article, you will learn how to create a basic Amazon Lex bot that can be used as a conversational bot with your configured chat service.

Please note that Bright Pattern only supports version 1 Amazon Lex bots at this time.

Procedure

This procedure will walk you through the process of setting up your first Amazon Lex bot. For a deeper understanding of Amazon Lex and other AWS resources, refer to AWS’s Amazon Lex Developer Guide.

Step 1: Create an AWS account

1. If you haven’t already done so, set up an AWS account.

Step 2: Create an IAM user, grant administrative permissions, and save credentials

Our integration accounts require access keys to connect to and use your Amazon Lex bot. Access keys are created and managed in AWS Identity and Access Management (IAM) services.

To get an access key, you need to:

1. Go to your AWS Management Console > IAM Dashboard. If you don't know where it is, search AWS Services for "IAM."
   
   
2. From the IAM root menu, click Users and then click Add user to create an IAM user and grant administrative permissions. Adding a user creates credentials that are used to access AWS.
   
   
   

   Add IAM User
   
   
   
3. Copy the access key ID (e.g., AKIAIOSFODNN7EXAMPLE) and the secret access key (e.g., wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY). Save this for when you set up an AWS Lex bot/chat suggestions integration account.
   
   

For more information, see Managing Access Keys for IAM Users.

Step 3: Add Lex as a service and create a sample bot

1. Sign in to the AWS Management Console and open the Amazon Lex console.
   
   
2. On the Create your bot page, you can create a custom bot or create a ready-to-use one with a sample template. For the sake of this example procedure, select the Book Trip sample.
   
   
3. The Create Your Bot page will open. That is where you develop your Amazon Lex bot.

Step 4: Review the bot workspace

The bot workspace will open. Notice that there are four tabs at the top: Editor, Settings, Channels, and Monitoring. In this exercise, we will be focusing on the Editor tab only. You can come back to the other tabs later.

The Editor tab includes the properties for every intent.

Step 5: Edit intent properties

Intents are actions triggered by keywords entered by your customer. You can think of intents as what customers want to do. For example, the first intent included in our sample Lex bot is "BookCar," which is what the customer wants to do (book a car) and what the bot recognizes it needs to do based on the customer's text input.

Properties

• Sample utterances - Phrases (i.e., keywords) that trigger the intent (e.g., "Reserve a car" or "make a car reservation" or "book a car")
  
  
• Lambda initialization and validation - AWS Lambda Function that validates the customer's input
  
  
• Slots - Data the customer must provide in order to fulfill the intent action (e.g., slot "PickUpCity" collects customer data about the city in which the car will be picked up)
  
  
• Confirmation prompt - Questions that confirm the intent actions and prompt the customer to input data (e.g., "Okay, I have you down for a car to be picked up in {PickUpCity}. Should I continue booking the car?")
  
  
• Fulfillment - AWS's business logic required to fulfill the customer's intent
  
  
  • AWS Lamda Function - The function used as a code hook for your Amazon Lex bot; this can perform initialization and validation, fulfillment, or both
    
    
  • Return parameters to client - Sends the intent information to the client application (i.e., Bright Pattern scenario) for intent fulfillment
    
    
• Response - Message(s) or response cards that close the intent or invoke a different one

Step 6: Save and build

1. For this example procedure, leave everything on the Editor tab as-is.
   
   
2. If you did change something, be sure to click Save at the bottom.
   
   
3. Click Build at the top of the page. This builds the bot with the configured intents.

Step 7: Test it

Once the build is complete, you can test the bot in the chat window.

1. On the right side of the screen, click Test chatbot to pop out the chat window.
   
   
2. Try typing a request such as, “I want to book a trip.” See what happens.

Step 8: Publish your bot

1. At the top of the page, click Publish.
   
   
2. In the Publish dialog that opens, choose or create a new alias (e.g., “TripBooker”) for this bot. The alias is used to point to the specific version of the bot. Having multiple aliases for the same bot allows you to keep and access different intents and properties for the same bot.
   
   
3. Click Publish again.

Next Steps

You have now set up a very basic Amazon Lex bot that can be integrated with Bright Pattern Contact Center. You may now:

• Read AWS documentation, explore Amazon Lex intents and slot types, and edit as desired
• Create an AWS Lex bot/chat suggestions integration account
• Edit a chat scenario that uses your bot
• Configure web chat to work with bots

Inbound Voice Service Configuration

This article describes how to configure inbound voice service within Bright Pattern’s Contact Center Administrator application.

For inbound voice services with virtual queuing (callback option), see also Bright Pattern's Virtual Queue Tutorial.

Prerequisites

This article assumes that you have already completed initial configuration for your contact center.

Procedure

1. Create and configure an inbound voice service

1. Log in to Contact Center Administrator, and go to Services and Campaigns.
   
   
2. Click the Add service (+) button to create a new inbound voice service.
   
   
3. Select Inbound Voice as the type of service.
   
   
   

   Choose from the available services
   
   
   
4. In the Properties tab that appears, update the service properties, being sure to give the service a unique name (required). At this point, all other properties are optional. For the sake of this procedure, we are going to keep the properties simple, as shown.
   
   
   

   Inbound voice service properties
   
   
   
5. Click Apply to save your changes.

2. Assign teams to the service

1. In the Assignments tab, assign teams to this service. Note that the service will be created with a default service skill that can be assigned automatically to all members of the associated teams.
   
   
   

   Assign users to teams
   
   
   
2. Optionally assign administrators and reviewers to this service.
   
   
3. Click Apply to save your changes.

3. Create additional skills and assign them

1. If necessary, create language skills and any other auxiliary skills that may have to be taken into account when routing calls requesting this service.
   
   
   

   You have to add skills before you can assign them
   
   
   
2. In Users & Teams > Skill Levels, use skill levels to assign the language and/or auxiliary skills to agents who will provide this service. The higher the numeric value, the more skilled that user is.
   
   
   

   Skills are assigned as percentage values (e.g., "100" or "50")
   
   
   
3. Click Apply to save your changes.

4. Create the scenario for the service

1. Go to Configuration > Scenarios > Chat to create the voice scenario for this service.
   
   
2. Either select an existing voice scenario from the list, or click the Add from template button at the bottom of the screen to create a new chat scenario from the “Inbound Voice” template.
   
   
   

   Select the "Inbound Voice" template
   
   
   
3. Creating a new scenario from a template will open the Scenario Builder application in a new browser tab or window. For the purpose of this simple setup, leave the scenario as-is and click Save. You can come back to Scenario Builder later to edit the scenario.
   
   
   

   Name the scenario
   
   
   
4. Give the scenario a unique name (e.g., "Inbound Sales") and click Save again. Your new scenario will appear in the list of scenarios.

5. Associate the scenario with a service

1. Go to Scenario Entries > Dial-in.
   
   
2. Select the dial-in scenario entry that has been preconfigured for your contact center (typically “Voicemail”).
   
   
3. In the Service property, select the inbound voice property that you created earlier in this procedure. This associates the scenario entry with the inbound voice service.
   
   
   

   Dial-in scenario entry properties
   
   
   
4. Click Apply to save your changes.

Note: The remaining steps are optional and depend on your workflows and reporting requirements.

6. Specify service level thresholds

1. Go back to Services & Campaigns and select the chat service you just created.
   
   
2. In the Service Level tab, set the percentage of chats associated with this service that are expected to be answered within a certain amount of time.
   
   
   

   Set service level thresholds

7. Specify dial-out information

1. Go to Directory > Dial-out Entries and select an entry from the list. A dial-out entry is an access number that has been assigned to your contact center by your service provider.
   
   
2. Update the entry properties, including phone number prefixes and caller ID, for outbound consultation calls.
   
   
   

   Update dial-out entries properties

8. Set up periodic call recording exports

1. Go to Services & Campaigns > Results tab.
   
   
2. Under Periodic Recording Export Jobs, click add.
   
   
3. Update all the properties, specifying which call recordings should be exported and where they should go.
   
   
   

   Periodic Recording Export Jobs properties

9. Configure caller ID

1. Go to Services & Campaigns and select your inbound voice service from the list of services.
   
   
2. Click on the Numbers tab and then on the Caller ID button to configure a caller ID for outbound SMS communications.
   
   
3. Click the Add caller ID (+) button.
   
   
4. Select your phone number from the list, and enter a state/province to be displayed.
   
   
5. Click Apply to save your changes.

10. Customize historical reports

Review the available service reports and, if necessary, customize them to your specific reporting needs.

Your inbound voice service configuration is now complete.

Service Configuration Quickstarts

Be sure to read Bright Pattern's other service configuration quickstarts, which will help you to set up bot-assisted chat services, campaigns, and email services for your contact center:

• Chat Service Configuration

• Outbound Campaign Service Configuration

• Email Service Configuration