Build Your First Use Case

Content on this page is for a product or feature in controlled release (CR). If you are not part of the CR group and would like more information, contact your CXone Account Representative.

Create Intents

Required permissionChatbot (ACD > DFO > Roles > [choose one] > Core Modules)

Intents are the core meaning behind the contact's words; the actual result or goal the contact wants to communicate or achieve. Follow these best practices when creating intents:

  • Use real data. Don't use tools that auto-generate data and claim to train your botClosed Virtual agent designed to handle specific interactions faster. They often produce examples that don't reflect what contacts really say. They also can result in a bot that loses its ability to generalize. Over time, the bot reaches a point where it only recognizes phrases it's seen before.
  • Don't use the same training examples for more than one intent. This may sound simple, but intents aren’t always so clear cut. Two user goals might seem different but start to gather similar examples over time. When this happens, see if you can merge similar intents into a more general category.
  • Always include an out-of-scope intent. Out-of-scope intents cover anything the user might say that’s outside of the bot’s domain. They allow you to recover the conversation and often result in improved performance.

Create a Single Intent

  1. In the Digital First Omnichannel portal, click Chatbot.
  2. Click the bot you want to work with.
  3. Click NLU NLU icon in the left icon menu.
  4. Click New Intent.
  5. Enter a name for your new intent and press Enter.
  6. Beneath the new intent, begin entering phrases that the bot should recognize for that intent. Separate phrases by pressing Enter. Entries are saved automatically.
  7. Click Train and Stage to train your bot on the new intent. You can also do more training and testing later.
  8. Once the intent is used in a story or rule, you can click View Dialogues, then click an item in that list to go to that story or rule.

Create a Multi-Intent

BotsClosed Virtual agent designed to handle specific interactions built with CXone Bot Builder can identify both single and multiple intentsClosed What the contact wants to communicate or accomplish in a contact's messageClosed Anything a contact says in a bot interaction, whether question or statement, written or spoken.. In the planning example, the final bot statement requires the bot to recognize two intents in the contact message.

Multi-intents can be as simple as the one in the planning example. They can also be more complicated and specific to your business. For example, a bank might need to create a multi-intent to handle a message like, "Please show me my balance and transfer $200 to my savings account."

If you create a multi-intent from two existing single intents, you don't have to provide many examples. Your bot can recognize the single intents in the message and handle them as a multi-intent.

Only use multi-intents when they are really necessary to the natural flow of conversation. Too many multi-intents can make your bot too complicated to manage easily.

  1. In the Digital First Omnichannel portal, click Chatbot.
  2. Click the bot you want to work with.
  3. Click NLU NLU icon in the left icon menu.
  4. Click New Intent.
  5. Enter a name for your new multi-intent and press Enter. The name should contain both intents connected by a + sign. You'll see Akela's multi-intent in the image: thanks + goodbye.
  6. Beneath the new multi-intent, begin entering phrases that the bot should recognize. Separate phrases by pressing Enter. Entries are saved automatically.
  7. Click Train and Stage to train your bot on the new multi-intent. You can also do more training and testing later.
  8. Once the multi-intent is used in a story or rule, you can click View Dialogues, then click an item in that list to go to that story or rule.

Create Rules

Required permissionChatbot (ACD > DFO > Roles > [choose one] > Core Modules)

Rules tell your botClosed Virtual agent designed to handle specific interactions how to respond to messagesClosed Anything a contact says in a bot interaction, whether question or statement, written or spoken. that are the same in every context. Examples include:

Conditions

You can define specific conditions that must be met in order for the rule to be used. You can choose from the following types of conditions:

  • Active Form—The rule only runs when the specific form or forms are active.
  • Slot Value—The rule only runs when the slot or slots have a specific value.

Add a New Rule

Before following these steps, be sure you create an intent for the rule, if you haven't already.

  1. In the Digital First Omnichannel portal, click Chatbot.
  2. Click the bot you want to work with.
  3. Click Dialogues Dialogues icon in the left icon menu.
  4. Click Rules.
  5. If your rule doesn't fit an existing folder, click New Folder, type a name for the new folder, and then press Enter.
  6. Under the folder you want to use, click New Rule.
  7. At the top of the rule section, enter a name for your rule.
  8. If your rule should only be used under a certain condition, click Add condition. Choose the condition type and the appropriate form or slot value.
  9. In the Customer may say box, enter a customer message and then press Enter. CXone Bot Builder analyzes the message and offers a list of intents and confidence level.
  10. Confirm the correct intent. Since your bot is learning a new intent, it may not display the correct one. You can select the correct intent from the drop-down.
  11. Hover over the plus sign at the left of the rule builder until it changes to a bot icon, then click the action.
  12. Provide the information matching the action type you selected. For example, if you selected Message, enter what the bot should say.

  13. Click the typing indicator to set or override the length of time a typing indicator displays for the action. If Smart Typing is turned on, the bot automatically determines this for you.
  14. You can add more examples to the intent. This helps ensure the rule is triggered for each way a customer may phrase something.
    • Click the intent under the contact message to load it in the Intents pane on the right. Then click Add example and add new messages.
    • Click the NLU icon , then click Add example for the intent you want add messages to.
  15. Click Train and Stage to train your bot on the new rule. You can also do more training and testing later.

Add New Messages to an Existing Rule

Before following these steps, be sure you create an intent for the rule, if you haven't already.

  1. In the Digital First Omnichannel portal, click Chatbot.
  2. Click the bot you want to work with.
  3. Click Dialogues Dialogues icon in the left icon menu.
  4. Click Rules.
  5. Click the rule you want to edit.
  6. Click the intent under the contact message to load it in the Intents pane on the right. Then click Add example and add new messages.
  7. Click Train and Stage to train your bot on the new rule. You can also do more training and testing later.

Create Stories

Required permissionChatbot (ACD > DFO > Roles > [choose one] > Core Modules)

Stories teach your botClosed Virtual agent designed to handle specific interactions how to respond to messagesClosed Anything a contact says in a bot interaction, whether question or statement, written or spoken. in the context of an interaction. They are typically expressed as happy and unhappy pathsClosed Story that produces a wrong outcome for the intent. Don't make up stories that you think might happen. Use real interactions to create them instead.

  1. In the Digital First Omnichannel portal, click Chatbot.
  2. Click the bot you want to work with.
  3. Click Dialogues Dialogues icon in the left icon menu.
  4. Click Stories.
  5. If the story doesn't fit in an existing folder, click New Folder, type a name for the new folder, and then press Enter.
  6. Under the new folder, click New Story.
  7. At the top of the story section, enter a name for your story.
  8. In the Customer may say box, enter the customer message and then press Enter. CXone Bot Builder analyzes the message and offers a list of intentsClosed What the contact wants to communicate or accomplish and confidence level.
  9. If CXone Bot Builder identified the correct intent, click Confirm. If not, expand the list of intents and select the correct intent, then click Confirm.
  10. Hover over the plus sign at the left of the rule builder until it changes to a bot icon, then click the action.
  11. Provide the information matching the action type you selected. For example, if you selected Message, enter what the bot should say.

  12. Click the typing indicator to set or override the length of time a typing indicator displays for the action. If Smart Typing is turned on, the bot automatically determines this for you.
  13. Repeat this process until you have completed the messages and actions for this story. If you need to edit intents or slots, you can do so right in the story builder.
  14. Click Train and Stage to train your bot on the new story. You can also do more training and testing later.

Design Fallback

Required permissionChatbot (ACD > DFO > Roles > [choose one] > Core Modules)

Fallback lets you tell a botClosed Virtual agent designed to handle specific interactions how to respond when it's not confident how to proceed. CXone Bot Builder includes two confidence thresholds for fallback design:

  • NLU Confidence Threshold—Measures how confident a bot is that it correctly identified a message's intentClosed What the contact wants to communicate or accomplish. The default level is 70%, so fallback is triggered at any level below that.
  • Action Confidence Threshold—Measures how confident a bot is about the next action it should take. The default level is 40%.
  1. In the Digital First Omnichannel portal, click Chatbot.
  2. Click the bot you want to work with.
  3. Click Dialogues Dialogues icon in the left icon menu.
  4. Click Fallback.
  5. If you want to use a custom value for NLU confidence threshold, enter that value in the field.
  6. If you want to use a custom value for Action confidence threshold, enter that value in the field.
  7. If you want to use a custom bot response when either threshold is met, enter your changes in the text field next to the bot icon.
  8. If you want the bot use the handoverClosed Any contact message that should trigger transfer to a live agent rule if it still doesn't understand, click the + icon and select Handover.
  9. Click Train and Stage to train your bot on the fallback design. You can also do more training and testing later.

Train Your Bot

You can duplicate the real-world interactions you collected earlier to make sure your bot handles them correctly. You can also create new interactions when you collect new training examples. As you're building your bot, the Train and Stage button lets you train your bot on new intentsClosed What the contact wants to communicate or accomplish, rulesClosed Used to define bot's response to messages that don't change with context., storiesClosed Used to train bot for interaction handling based on intent and context, and so on.

Talk to Your Bot

  1. In the Digital First Omnichannel portal, click Chatbot.
  2. Click the bot you want to work with.
  3. Click Learning Learning icon in the left icon menu.
  4. Click Talk to your bot.
  5. Enter a messageClosed Anything a contact says in a bot interaction, whether question or statement, written or spoken. in the field at the bottom of the chat window, just as you would in any chat conversation. Your bot responds as it would in a real interaction.
  6. Continue to enter messages and review the bot's responses. If at any point the bot responds in an unexpected way, you can stop and correct the issue. For example, you might need to create a new intentClosed What the contact wants to communicate or accomplish or ruleClosed Used to define bot's response to messages that don't change with context..
  7. If you need to restart the interaction after correcting an issue, click Start New Conversation.