Getting Started with Agent Builder

This page introduces essential Conversational AI concepts and ties them to Agent Builder. The goal is to help you understand the purpose of Agent Builder configurations and what they help you accomplish.

A conversation between a contact and any agent, including Mpower Agents A virtual agent created with CXone Mpower Agent Builder that can handle voice or chat interactions. or other virtual agents A software application that handles customer interactions in place of a live human agent., has three key elements. The key conversational elements have corresponding configurations in Agent Builder. When you understand these configurations and how they work together, you can approach creating an Mpower Agent with confidence. They are:

What the contact says.
What the contact wants.
What the agent or Mpower Agent says and does.

What the contact says and wants are represented in Agent Builder with intents, entities, and slots. What the Mpower Agent says and does is represented by stories, rules, and Mpower Agent actions.

What the Contact Says and Wants: Utterances, Intents, Entities, and Slots

The contact The person interacting with an agent, IVR, or bot in your contact center. communicates with the Mpower Agent A virtual agent created with CXone Mpower Agent Builder that can handle voice or chat interactions. by typing messages in the chat window or by speaking into the phone. If the interaction is a phone call, audio captured from the call is transcribed to text using a speech-to-text Also called STT, this process converts spoken language to text. service. The Mpower Agent receives the messages or transcribed audio—also called utterances—and analyzes them. Then, the Mpower Agent can act on them based on how it's configured to respond.

Concept	Definition	Example	What the Mpower Agent Does
Utterance	Anything a contact The person interacting with an agent, IVR, or bot in your contact center. says in an interaction The full conversation with an agent through a channel. For example, an interaction can be a voice call, email, chat, or social media conversation.. Sometimes called a message.	"I lost my password." "What is my balance?" "Are you a bot?"	The Mpower Agent uses Natural Language Understanding (NLU) to analyze each contact utterance to determine its meaning, or intent.
Intent	What the contact wants to communicate or accomplish. Every message the contact sends has an intent.	"I lost my password" has the intent of "reset password". "Hello" has the intent of "greeting".	The Mpower Agent analyzes a contact's message using NLU This process expands on Natural Language Processing (NLP) to make decisions or take action based on what it understands. to determine the intent. Once it knows that, it can respond with a message of its own. You configure the response you want your Mpower Agent to use for each intent.
Entity	A defined piece of information in a contact's message.	Person or product name, phone number, account number, location, and so on.	The Mpower Agent uses NLU to identify entities in a contact's message. Entities help the Mpower Agent understand what the contact's message means.
Slot	An entity extracted from a contact's message and saved for use in Mpower Agent responses. Similar to a variable.	Creating a slot for contact name lets the Mpower Agent use that name in responses during an interaction, making it more personal.	When configured to do so, the Mpower Agent extracts an entity from a contact message and saves it in a slot. You can have your Mpower Agent use this information later in the conversation.

What the Mpower Agent Says and Does: Stories, Rules, and Actions

Human conversation is unpredictable and varied. Mpower Agent A virtual agent created with CXone Mpower Agent Builder that can handle voice or chat interactions. responses are not. This means your Mpower Agent must be able to correctly interpret the wide variation in how humans speak, but it doesn't have to "think" about how it will respond. Mpower Agent responses are clearly defined in the Agent Builder configurations. However, whether the Mpower Agent chooses the correct response for the situation depends on how well it identifies the intent The meaning or purpose behind what a contact says/types; what the contact wants to communicate or accomplish. for each contact utterance What a contact says or types..

Concept	Definition	Example	What the Mpower Agent Does
Rule	Defines Mpower Agent responses to messages that don't change meaning with context.	Single-turn interactions with fixed responses: What are your hours? What is your address? Conversation building blocks: Greetings, good-byes, thank yous and transitions; simple yes/no questions; and acknowledgments. Mpower Agents come with default intents and rules for several of these, including greetings, handover The transfer of a contact from a virtual agent to a live agent. requests, and more. FAQs Insults and classic bot challenges	Rules are one of two ways you can configure how your Mpower Agent responds to an intent. Rules are useful for certain kinds of intents, but not all intents.
Story	Trains an Mpower Agent to handle an interaction based on message intent and conversational context.	In an interaction about a forgotten password, the Mpower Agent would respond to "How do I do that?" in one way. If the interaction were about creating a new account, the response would be quite different even though in both cases the contact is using the same words with the same intent—to get more information.	Stories are the second of two ways you can configure how your Mpower Agent responds to an intent. Stories teach the Mpower Agent how to use the context of the conversation to respond appropriately.
Mpower Agent Action	Anything an Mpower Agent says or does while handling an interaction.	In an interaction about a forgotten password, the Mpower Agent responds by sending the link to the password reset FAQ on the website. When a contact expresses frustration, such as "I don't understand! It's not working!!!" the Mpower Agent responds with "I'm sorry. Would you like me to transfer you to a human agent?" When the contact says yes, the Mpower Agent initiates the transfer.	Mpower Agent actions are the options you have when defining how you want your Mpower Agent to respond to each intent. They give you the flexibility to configure each response to achieve the outcome that meets the contact's needs.

Concept

Definition

Example

What the Mpower Agent Does

Rule

Defines Mpower Agent responses to messages that don't change meaning with context.

Single-turn interactions with fixed responses: What are your hours? What is your address?
Conversation building blocks: Greetings, good-byes, thank yous and transitions; simple yes/no questions; and acknowledgments. Mpower Agents come with default intents and rules for several of these, including greetings, handover The transfer of a contact from a virtual agent to a live agent. requests, and more.
FAQs
Insults and classic bot challenges

Rules are one of two ways you can configure how your Mpower Agent responds to an intent. Rules are useful for certain kinds of intents, but not all intents.

Story

Trains an Mpower Agent to handle an interaction based on message intent and conversational context.

In an interaction about a forgotten password, the Mpower Agent would respond to "How do I do that?" in one way. If the interaction were about creating a new account, the response would be quite different even though in both cases the contact is using the same words with the same intent—to get more information.

Stories are the second of two ways you can configure how your Mpower Agent responds to an intent. Stories teach the Mpower Agent how to use the context of the conversation to respond appropriately.

Mpower Agent Action

Anything an Mpower Agent says or does while handling an interaction.

In an interaction about a forgotten password, the Mpower Agent responds by sending the link to the password reset FAQ on the website.

When a contact expresses frustration, such as "I don't understand! It's not working!!!" the Mpower Agent responds with "I'm sorry. Would you like me to transfer you to a human agent?" When the contact says yes, the Mpower Agent initiates the transfer.

Mpower Agent actions are the options you have when defining how you want your Mpower Agent to respond to each intent. They give you the flexibility to configure each response to achieve the outcome that meets the contact's needs.

How the Mpower Agent Learns: Training and Testing

Training teaches your Mpower Agent to correctly predict contacts' intents The meaning or purpose behind what a contact says/types; what the contact wants to communicate or accomplish.. It starts with providing your Mpower Agent plenty of high-quality, real-world examples of each intent. This is called training data. When combined with training dialogues Mpower Agent stories, rules, and flows in Agent Builder., training data helps your Mpower Agent learn to recognize what contacts The person interacting with an agent, IVR, or bot in your contact center. need, and then to respond appropriately.

To test your Mpower Agent, you must first build a model Version of a bot that has been trained and staged of the data in the its configurations. In Agent Builder, the Train and Stage option triggers a new model to be built. The process runs in the background.

A model is built from an analysis of the configured intents, stories, rules, examples, and other training data. The model is the core of what your Mpower Agent is—a software program that analyzes conversational human speech to determine the closest match from its set of data points, then executes the corresponding action.

You can build a new model as often as you want to. Each model is numbered and there is a historical list of past models.

After the model is built, you can test your Mpower Agent. Testing involves having conversations with your Mpower Agent. This is how you find the places where it doesn't understand your messages or makes poor predictions. You can fix the configurations, build a new model, and test again.

Agent Builder has a built-in chat program you can use to have test conversations with your Mpower Agent. You can also share your Mpower Agent with other people who can help you test it. They don't have to report their experiences to you, because every conversation your Mpower Agent has is saved in Agent Builder. You can review them to find the pain points that need attention.

Reviewing conversation data is important during the development of your Mpower Agent as well as after it's in production. Ongoing fine-tuning of intents, stories, and rules is needed to ensure your Mpower Agentcontinues to perform well.

How to Deal with Trouble

Agent Builder provides two configurations that allow you to deal with potential Mpower Agent problems:

Fallback: This teaches your Mpower Agent what to do when it's not sure how to proceed. There are two kinds of fallback:
- NLU fallback: Used when the Mpower Agentisn't confident its understanding the contact.
- Action fallback: Used when the Mpower Agent isn't confident in its ability to predict the next action.
Safety Net: A safety net allows you to configure what happens when there's another problem with the Mpower Agent or the systems it connects to. This could include things such as the Mpower Agent taking longer than normal to respond to the contact.

Fallback and a safety net are preventative options, but not all problems can be prevented. It's important to regularly review conversation data to look for contact pain points.

Monitor and Manage Your Mpower Agents

Agent Builder provides many tools you can use to monitor the performance of your Mpower Agent. This is a critical, ongoing part of managing your Mpower Agents. By regularly monitoring their performance with these tools, you can spot pain points and refine the Mpower Agent configurations to ease them.

The following tools allow you to review conversation data:

Insights: Provides reporting and real-time, interactive analytics for your Mpower Agents:
NLU Inbox: Helps you manage your NLU This process expands on Natural Language Processing (NLP) to make decisions or take action based on what it understands. data to improve the quality of your Mpower Agent. It shows all new messages from contacts.
Query search: Use the tags to narrow results in the NLU inbox or the Insights section.

The following tools allow you to manage and organize your Mpower Agent data:

Tags: Use tags throughout Agent Builder. You can have them applied automatically by the Mpower Agent or you can apply them manually.
Bot skills: Use bot skills to organize training data based on what your Mpower Agent can do. You can filter training data by skill to limit what's visible to you.

The following tools allow you to view information about your Mpower Agent:

Health Monitor: Displays information about the training, models, and configuration changes for your Mpower Agent.
Import and Export Tools: Import or export certain data from your Mpower Agents. You can use this as a backup option.
Activity Log: Provides a history of what users are doing when logged in to Agent Builder.

Getting the Most from Agent Builder

As you begin to make plans for how best to work Mpower Agents into your contact center, consider the following ideas. They can help you get the most from Agent Builder.

Build multiple Mpower Agents for different use cases, channels, or audiences. You can have them work together with live human agents using digital Any channel, contact, or skill associated with Digital Experience.ACD skills. CXone Mpower views your Mpower Agent as a user entity, so routing works the same way for them as for your human agents.
There are many use cases for Mpower Agents. For example, you can:
- Collect information before handing the contact over to a human agent.
- Handle triage at the start of interactions to route contacts to a more specific agent.
- Use bots to handle the most common and simple questions your agents receive, such as checking on order statuses or billing due dates, updating contact information, or questions about store locations and hours.
- Let bots handle your night shift to provide 24/7 customer service. Create a digital ACD skill for overnight interactions, then set up your script with to send incoming interactions to a Mpower Agent overnight. The Mpower Agent can either handle the interaction or route it to an agent for handling the next morning.
- Use Autopilot Knowledge to deliver answers pulled directly from your Expert knowledge management center.
- Set up integrations between Agent Builder and other systems via API to increase the usefulness of your bot.
- Set up scripting integrations to create custom Mpower Agent actions using JavaScript.
- Use them to carry out tasks for your live agents. Your live agents must use CXone Mpower Copilot for Agents to use Mpower Agents this way.

How to Start a New Mpower Agent Project

If you're new to Agent Builder, you can follow the tutorial. It walks you through creating a sample Mpower Agent. It also includes steps you can follow in Agent Builder to create your first Mpower Agent.

After you are comfortable with the Agent Builder interface and concepts, you can follow the implementation process to begin planning and building a Mpower Agent. This process covers all the tasks you need to complete to create and manage your Mpower Agents.

After you have a stable working model of your Mpower Agent, you can begin to slowly introduce it to your customers. Instead of immediately having your Mpower Agent handle a full load of interactions from the start, you could:

Drive a small percentage of traffic to your Mpower Agent to start out slowly. Raise the percentage over time as it gets smarter and more capable of handling more use cases. In your Studio script, use script logic to set conditions to define what traffic should be routed to your Mpower Agent.
Use digital ACD skills Used to automate delivery of interactions based on agent skills, abilities, and knowledge. and script logic to split traffic between two different Mpower Agents for A/B testing. This helps you validate which Mpower Agent performs better.