Skip to content

LLM Based Conversation Summary

Overview

Contact center administrators can integrate external Gen AI models to enhance conversation summarization capabilities. This feature enables you to:

  • Use custom models for conversation summarization
  • Support additional languages not natively available in CCAI
  • Create customized summarization prompts

Prerequisites

Before configuring external Gen AI models, ensure you have:

  • Access to the XO11 platform
  • Administrative privileges
  • Gen AI connector provided by the platform team
  • A configured external Gen AI model

Enabling Gen AI-Based Summarization

  1. Go to Contact Center > CONFIGURATIONS > Advanced settings > LLM based Conversation Summary.
    LLM Based Conversation Summary

  2. Turn on the LLM based Conversation Summary toggle.

  3. Select your preferred summarization method:

    • OOTB (Out-of-the-box) Kore Summarisation (Default)
    • Using the model configured inside the GenAI framework

      Conversation Summary Options

    Note

    The GenAI framework option remains disabled until you configure a valid model.

Configuring a Custom Model

To set up a custom summarization model:

  1. Access the GenAI & LLM on the XO11 Platform.
  2. Configure your external Gen AI model. Learn more.
  3. Define custom prompts for your specific summarization needs.

    Note

    If you try to use the GenAI framework without a configured model, you'll receive the message: "A custom model and an appropriate prompt need to be built from inside the Gen AI tools."

How Summarization Works

Once configured, the summarization process:

  1. Captures the conversation transcript.
  2. Sends the transcript to your configured Gen AI model.
  3. Applies your custom prompt.
  4. Displays the summarized conversation to your agents.

Benefits

Contact Center Administrators can:

  • Configure external models and custom prompts.
  • Monitor summarization across additional languages.
  • Extend language support beyond native capabilities.

Agents gain:

  • Enhanced summarization flexibility.
  • Broader language support.
  • Customized summary outputs based on defined prompts.