AgentAssist Widget

The AgentAssist is an AI powered virtual assistant to optimize contact center agent performance. It offers seamless assistance, promoting enhanced productivity and improved customer satisfaction. Its user-friendly interface and real-time support capabilities let agents effortlessly solve complex customer interactions.

The following features are available in the AgentAssist widget:

  • Dialog Tasks: Configured Use Cases that trigger when a customer’s utterance matches the content of a Dialog Task. Designed to help agents minimize manual typing by copy-pasting the relevant answers from the Dialog Tasks. Learn more.
  • Playbooks: Contact center agents are expected to provide the best customer experience despite high call volumes and complex interactions. To empower agents in this endeavor, AgentAssist introduces the Playbook feature—a standard tool offering structured guidance for handling customer interactions. Learn more.
  • Real-time Agent Coaching: Real-time coaching for the agents based on the analysis of agent utterances and speech. You can set up specific utterances and speech patterns for AgentAssist to take relevant actions, such as providing hints and nudges or emailing the manager when those patterns occur. Learn more.
  • Real-time Sentiment Analysis: Provides a continuous, time-based graph that displays the customer’s sentiment fluctuations. Agents get immediate visibility of the evolving emotional tone of the conversation and take proactive and empathetic responses. Learn more.
  • Transcription: Enhances agent accessibility and context by providing accurate transcriptions of user and agent voices during interactions. Offers a conversation overview and serves as the basis for automated Quality Assessment (QA). Agents receive a visual indicator when an intent is identified, improving real-time understanding and response.
  • Auto Summarization: Automatically shortens lengthy information into concise summaries, enabling agents to respond promptly to customer inquiries. It helps agents quickly understand and make decisions.
  • PII Redaction: Securely removes Personally Identifiable Information (PII) from the agent’s view and transcripts, ensuring a safe environment for customer interactions and compliance with privacy regulations. Learn more.
  • Feedback: Empowers agents to provide real-time feedback on suggestions. This continuous process forms a dynamic loop, allowing Machine Learning (ML) algorithms to continuously enhance and customize recommendations for accurate and contextually relevant outcomes.
  • Next Best Action: Enhances the product by enabling seamless execution of Dialog Tasks for swift customer query resolution. The generative AI nodes and script nodes integration equips the system to efficiently communicate with external systems through API calls. This leads to improved customer experience and reduced necessity for frequent screen changes for agents. This allows agents to deliver contextually relevant next-best responses, contributing to a more streamlined and effective contact center operation.
  • Multi-language Support: Provides multiple language support, enabling agents to switch between languages instantaneously. It ensures seamless communication with users in diverse linguistic environments. Agents can engage with customers in their preferred language, providing a more personalized and efficient support experience.
  • Knowledge AI: Allows customization of Knowledge AI search result displays with control over search results presentation, the option to display auto suggestions in the Ask AgentAssist search bar, and optimization of the search experience through advanced Knowledge AI API configuration, offering both Basic and Advanced options.

Widget Tabs

The widget consists of five tabs–Search, My Bot, Assist, Transcript, and Settings–to help the agents perform different tasks and activities.

AgentAssist tabs

Search Tab

This tab shows all configured Dialog Tasks at the bot level and also gives search capabilities wherein agents can search Dialog Tasks, frequently asked questions (FAQs), Generative AI answers, web pages, files, and more. The content in this section dynamically updates based on all agent interactions. The agents can scroll through the list to get visibility of all the automations that are readily available to execute. If you click the ellipsis and select “Run with Agent’s input”, the dialog moves to the My Bot section, and you can run the dialog from there.

The following buttons are available in this section:

  • Run button: For Dialog Tasks. Click to run a configured Dialog Task.
  • Run with Agent’s input: Click this button to run a Dialog Task with agent’s input. The agent can customize the content and decide whether to send a dialog or not. Selecting this option automatically moves you to the My Bot section.

My Bot Tab

My Bot functionality allows agents to run the entire automation by themselves. Agents provide inputs to the bot and obtain the desired results.

My Bot tab displays the list of “next best action” suggestions for the agent based on the current conversation status and other interaction parameters.

My bot functionality can be used in multiple circumstances (but not limited to): 

  • If Automation fails, the agent can rerun it by providing inputs manually, which helps to reduce the average handle time (AHT) and makes it less frustrating for customers as they don’t have to repeat the same information.
  • Agents can run MyBot automation in parallel and seek information directly from the bot.

Assist Tab

The Assist tab is the primary interface for the agents. AgentAssist actively monitors conversations and offers assistance to agents. Through the Assist tab, customer intents in the form of Dialog Tasks, FAQs, KnowledgeAI, and Playbook are identified. Agents can run automation processes and send FAQs, Playbook body text, and Dialog Task text to users. The Assist tab displays Dialog Tasks, Playbooks, and Agent Coaching progression. For these features to be available, they should be configured on your device. This section also displays a “bulb” icon if there is a Dynamic Playbook match.

The Assist tab facilitates the following:

  • Guide agents through conversations by providing the best response suggestions.
  • Automatic retrieval of information from knowledge base (KB) based on user intent.
  • Provide structured training and coaching to agents.
  • Run bot for customers based on customer-provided information.
  • Provide a historical view to look at the previously executed automation.
  • View Conversation History for chat conversations– The History tab shows the record of user-bot interaction.
  • Send/Copy button 
  • Feedback for Dialog Task and Knowledge AI

The Assist tab features the following:

  • Running Playbook stages: Shows the total and adhered stages of a running playbook. For example, 1/3 means 1 out of 3 stages of a playbook has been adhered to.
  • Proceed to Close: This button is available when a playbook is running. Clicking this button leads to a change in the playbook running status from Open to Close, and agents are prompted to ask if the customer needs any other help before ending the call.
  • Playbook adherence button: If a Playbook is running, adherence to a step happens automatically when the agent’s utterance matches exactly with the Playbook step adherence utterance. If it doesn’t match, the agent can click this button to indicate a step has been completed (adhered).
  • Send/Copy buttons: These two buttons appear for Playbooks (if enabled) and Dialog Tasks. Click Send to send the exact message appearing on the widget, or click Copy to customize the message before sending it to the customer.
  • Resume button: This button applies to the Playbooks and appears while switching to another playbook while one is already running.
  • List View: This is an entity list button. When clicked, a pop-up displays a list of all entity entries present in the conversation.
  • Restart button: When clicked, this button provides agents with two options: “Start over” and “Restart with Captured Inputs.” These options allow agents to either restart the Dialog Task or continue it by providing input for each entity before restarting.
  • Termination button: This button provides agents options to handle ongoing interactions effectively, allowing them to either terminate, cancel, or continue, which involves manual input of customer responses for entity responses.
    • Option 1: Yes, Terminate: Terminates the running Dialog Task and provides feedback.
    • Option 2: No, Cancel: Closes the pop-up with no actions taken.
    • Option 3: Continue with Override Mode: Agents manually enter the response to the customer’s conversation.
  • Maximize/Minimize button: This button maximizes/minimizes the displayed page section.

Transcript Tab

The Transcript feature is available specifically for voice calls.

When a voice call begins, the Transcript tab shows the transcripts. This information is in text format, which allows the agent to look up information and copy/paste it.

The Transcript functionality transcribes the audio for the agent and the customer in real-time. Once transcribed, the messages are readily available for the agent throughout the session. Also, it allows the agent to have a historical view for referring back whenever needed.

Settings

The Settings tab gives you access to set the listening mode of the bot, available documents and FAQs, and the option to choose a language.

Proactive Mode

The Proactive Mode toggle allows you to put the bot in listening mode. You can turn off the toggle if it is not working as expected.

Help & Support

This section gives you access to the available documentation, FAQs, and the Kore Academy courses. You can refer to these materials to answer questions or give reference to these to the customer.

Language

This section lists the supported languages and lets you select a language of your choice. The selected language appears at the top of the section heading.

AgentAssist V2 vs V3 – Key Differences

AgentAssist V3 introduces many additional features and updates over AgentAssist V2. This document highlights the differences visible on the user interface (UI) and at the feature-level.

Difference in the UI

The following UI updates are done:

  • Tab Configuration:
    • V2 features three tabs – Assist, Library, and My Bot – and a search bar at the bottom.
    • V3 features four tabs – Search, My Bot, Assist, and Settings.
  • Library Tab Transformation: The Library tab in V2 has been transformed into the Search tab in V3.
  • Tab Placement: In V2, tabs are positioned at the top but in V3 they are moved to the bottom of the screen.
  • Sentiment Analysis Display: The Sentiment Analysis display is always visible in V3, but it was visible only upon a customer utterance match in V2. V3 has badges, an icon (for example, Positive, Negative and Neutral), and an emoji for every customer utterance. Clicking the badge takes you to the sentiment graph (as was the case in V2).
  • Proactive Button Location: The Proactive button, earlier located at the top-right corner of the page in V2 has been moved to the Settings tab in V3.
  • Playbook Closure: The Proceed to Close display for Playbooks in V2 has been replaced with a symbol in V3.
    • The Widget Playbook list in V2 displays only the Primary Playbooks and doesn’t display the Dynamic Playbooks.
    • The Widget Playbook list in V3 displays both the Primary and the Dynamic Playbooks.
  • Dialog Task Section: The Dialog Task section in V2 has the customer’s utterance displayed at the top, but it is updated as a tool tip (accessible with a hover over the information icon) in V3.
    • Additional Information in V3: The Dialog Task section in V3 has additional information–Wrong Suggestions? feedback button.

AgentAssist V2

AgentAssist V3

 

Differences at the Feature/Tab level

The following updates have been added at the feature level:

  • Dialog Task execution: The Dialog Tasks in V2 had Terminate and Override Input buttons, but V3 has List View, Restart, Terminate, and Maximize functionality as symbols.

    AgentAssist V2
    AgentAssist V3
  • The Override Input button in V2 and the input field under the running Dialog Task in V3 have the same functionality – prevent the widget from capturing the customer’s input. This functionality is useful when the agent has multiple conversations going on simultaneously.

    AgentAssist V2

    AgentAssist V3

New Tab addition – Settings Tab

The Settings tab in V3 lets you to:

AgentAssist Widget

The AgentAssist is an AI powered virtual assistant to optimize contact center agent performance. It offers seamless assistance, promoting enhanced productivity and improved customer satisfaction. Its user-friendly interface and real-time support capabilities let agents effortlessly solve complex customer interactions.

The following features are available in the AgentAssist widget:

  • Dialog Tasks: Configured Use Cases that trigger when a customer’s utterance matches the content of a Dialog Task. Designed to help agents minimize manual typing by copy-pasting the relevant answers from the Dialog Tasks. Learn more.
  • Playbooks: Contact center agents are expected to provide the best customer experience despite high call volumes and complex interactions. To empower agents in this endeavor, AgentAssist introduces the Playbook feature—a standard tool offering structured guidance for handling customer interactions. Learn more.
  • Real-time Agent Coaching: Real-time coaching for the agents based on the analysis of agent utterances and speech. You can set up specific utterances and speech patterns for AgentAssist to take relevant actions, such as providing hints and nudges or emailing the manager when those patterns occur. Learn more.
  • Real-time Sentiment Analysis: Provides a continuous, time-based graph that displays the customer’s sentiment fluctuations. Agents get immediate visibility of the evolving emotional tone of the conversation and take proactive and empathetic responses. Learn more.
  • Transcription: Enhances agent accessibility and context by providing accurate transcriptions of user and agent voices during interactions. Offers a conversation overview and serves as the basis for automated Quality Assessment (QA). Agents receive a visual indicator when an intent is identified, improving real-time understanding and response.
  • Auto Summarization: Automatically shortens lengthy information into concise summaries, enabling agents to respond promptly to customer inquiries. It helps agents quickly understand and make decisions.
  • PII Redaction: Securely removes Personally Identifiable Information (PII) from the agent’s view and transcripts, ensuring a safe environment for customer interactions and compliance with privacy regulations. Learn more.
  • Feedback: Empowers agents to provide real-time feedback on suggestions. This continuous process forms a dynamic loop, allowing Machine Learning (ML) algorithms to continuously enhance and customize recommendations for accurate and contextually relevant outcomes.
  • Next Best Action: Enhances the product by enabling seamless execution of Dialog Tasks for swift customer query resolution. The generative AI nodes and script nodes integration equips the system to efficiently communicate with external systems through API calls. This leads to improved customer experience and reduced necessity for frequent screen changes for agents. This allows agents to deliver contextually relevant next-best responses, contributing to a more streamlined and effective contact center operation.
  • Multi-language Support: Provides multiple language support, enabling agents to switch between languages instantaneously. It ensures seamless communication with users in diverse linguistic environments. Agents can engage with customers in their preferred language, providing a more personalized and efficient support experience.
  • Knowledge AI: Allows customization of Knowledge AI search result displays with control over search results presentation, the option to display auto suggestions in the Ask AgentAssist search bar, and optimization of the search experience through advanced Knowledge AI API configuration, offering both Basic and Advanced options.

Widget Tabs

The widget consists of five tabs–Search, My Bot, Assist, Transcript, and Settings–to help the agents perform different tasks and activities.

AgentAssist tabs

Search Tab

This tab shows all configured Dialog Tasks at the bot level and also gives search capabilities wherein agents can search Dialog Tasks, frequently asked questions (FAQs), Generative AI answers, web pages, files, and more. The content in this section dynamically updates based on all agent interactions. The agents can scroll through the list to get visibility of all the automations that are readily available to execute. If you click the ellipsis and select “Run with Agent’s input”, the dialog moves to the My Bot section, and you can run the dialog from there.

The following buttons are available in this section:

  • Run button: For Dialog Tasks. Click to run a configured Dialog Task.
  • Run with Agent’s input: Click this button to run a Dialog Task with agent’s input. The agent can customize the content and decide whether to send a dialog or not. Selecting this option automatically moves you to the My Bot section.

My Bot Tab

My Bot functionality allows agents to run the entire automation by themselves. Agents provide inputs to the bot and obtain the desired results.

My Bot tab displays the list of “next best action” suggestions for the agent based on the current conversation status and other interaction parameters.

My bot functionality can be used in multiple circumstances (but not limited to): 

  • If Automation fails, the agent can rerun it by providing inputs manually, which helps to reduce the average handle time (AHT) and makes it less frustrating for customers as they don’t have to repeat the same information.
  • Agents can run MyBot automation in parallel and seek information directly from the bot.

Assist Tab

The Assist tab is the primary interface for the agents. AgentAssist actively monitors conversations and offers assistance to agents. Through the Assist tab, customer intents in the form of Dialog Tasks, FAQs, KnowledgeAI, and Playbook are identified. Agents can run automation processes and send FAQs, Playbook body text, and Dialog Task text to users. The Assist tab displays Dialog Tasks, Playbooks, and Agent Coaching progression. For these features to be available, they should be configured on your device. This section also displays a “bulb” icon if there is a Dynamic Playbook match.

The Assist tab facilitates the following:

  • Guide agents through conversations by providing the best response suggestions.
  • Automatic retrieval of information from knowledge base (KB) based on user intent.
  • Provide structured training and coaching to agents.
  • Run bot for customers based on customer-provided information.
  • Provide a historical view to look at the previously executed automation.
  • View Conversation History for chat conversations– The History tab shows the record of user-bot interaction.
  • Send/Copy button 
  • Feedback for Dialog Task and Knowledge AI

The Assist tab features the following:

  • Running Playbook stages: Shows the total and adhered stages of a running playbook. For example, 1/3 means 1 out of 3 stages of a playbook has been adhered to.
  • Proceed to Close: This button is available when a playbook is running. Clicking this button leads to a change in the playbook running status from Open to Close, and agents are prompted to ask if the customer needs any other help before ending the call.
  • Playbook adherence button: If a Playbook is running, adherence to a step happens automatically when the agent’s utterance matches exactly with the Playbook step adherence utterance. If it doesn’t match, the agent can click this button to indicate a step has been completed (adhered).
  • Send/Copy buttons: These two buttons appear for Playbooks (if enabled) and Dialog Tasks. Click Send to send the exact message appearing on the widget, or click Copy to customize the message before sending it to the customer.
  • Resume button: This button applies to the Playbooks and appears while switching to another playbook while one is already running.
  • List View: This is an entity list button. When clicked, a pop-up displays a list of all entity entries present in the conversation.
  • Restart button: When clicked, this button provides agents with two options: “Start over” and “Restart with Captured Inputs.” These options allow agents to either restart the Dialog Task or continue it by providing input for each entity before restarting.
  • Termination button: This button provides agents options to handle ongoing interactions effectively, allowing them to either terminate, cancel, or continue, which involves manual input of customer responses for entity responses.
    • Option 1: Yes, Terminate: Terminates the running Dialog Task and provides feedback.
    • Option 2: No, Cancel: Closes the pop-up with no actions taken.
    • Option 3: Continue with Override Mode: Agents manually enter the response to the customer’s conversation.
  • Maximize/Minimize button: This button maximizes/minimizes the displayed page section.

Transcript Tab

The Transcript feature is available specifically for voice calls.

When a voice call begins, the Transcript tab shows the transcripts. This information is in text format, which allows the agent to look up information and copy/paste it.

The Transcript functionality transcribes the audio for the agent and the customer in real-time. Once transcribed, the messages are readily available for the agent throughout the session. Also, it allows the agent to have a historical view for referring back whenever needed.

Settings

The Settings tab gives you access to set the listening mode of the bot, available documents and FAQs, and the option to choose a language.

Proactive Mode

The Proactive Mode toggle allows you to put the bot in listening mode. You can turn off the toggle if it is not working as expected.

Help & Support

This section gives you access to the available documentation, FAQs, and the Kore Academy courses. You can refer to these materials to answer questions or give reference to these to the customer.

Language

This section lists the supported languages and lets you select a language of your choice. The selected language appears at the top of the section heading.

AgentAssist V2 vs V3 – Key Differences

AgentAssist V3 introduces many additional features and updates over AgentAssist V2. This document highlights the differences visible on the user interface (UI) and at the feature-level.

Difference in the UI

The following UI updates are done:

  • Tab Configuration:
    • V2 features three tabs – Assist, Library, and My Bot – and a search bar at the bottom.
    • V3 features four tabs – Search, My Bot, Assist, and Settings.
  • Library Tab Transformation: The Library tab in V2 has been transformed into the Search tab in V3.
  • Tab Placement: In V2, tabs are positioned at the top but in V3 they are moved to the bottom of the screen.
  • Sentiment Analysis Display: The Sentiment Analysis display is always visible in V3, but it was visible only upon a customer utterance match in V2. V3 has badges, an icon (for example, Positive, Negative and Neutral), and an emoji for every customer utterance. Clicking the badge takes you to the sentiment graph (as was the case in V2).
  • Proactive Button Location: The Proactive button, earlier located at the top-right corner of the page in V2 has been moved to the Settings tab in V3.
  • Playbook Closure: The Proceed to Close display for Playbooks in V2 has been replaced with a symbol in V3.
    • The Widget Playbook list in V2 displays only the Primary Playbooks and doesn’t display the Dynamic Playbooks.
    • The Widget Playbook list in V3 displays both the Primary and the Dynamic Playbooks.
  • Dialog Task Section: The Dialog Task section in V2 has the customer’s utterance displayed at the top, but it is updated as a tool tip (accessible with a hover over the information icon) in V3.
    • Additional Information in V3: The Dialog Task section in V3 has additional information–Wrong Suggestions? feedback button.

AgentAssist V2

AgentAssist V3

 

Differences at the Feature/Tab level

The following updates have been added at the feature level:

  • Dialog Task execution: The Dialog Tasks in V2 had Terminate and Override Input buttons, but V3 has List View, Restart, Terminate, and Maximize functionality as symbols.

    AgentAssist V2
    AgentAssist V3
  • The Override Input button in V2 and the input field under the running Dialog Task in V3 have the same functionality – prevent the widget from capturing the customer’s input. This functionality is useful when the agent has multiple conversations going on simultaneously.

    AgentAssist V2

    AgentAssist V3

New Tab addition – Settings Tab

The Settings tab in V3 lets you to: