Creating a Gen AI Voice Bot

A Gen AI Voice Bot enables automated voice interactions powered by generative AI. Using Chat360’s Voice Bot Builder, you can create conversational bots that understand user input through Speech-to-Text (STT), process queries using Large Language Models (LLMs), and respond through Text-to-Speech (TTS).

This guide explains how to create and configure a Gen AI-powered voice bot from login to deployment.

Step 1: Log In to the Platform

• Open your web browser.

• Navigate to the voice bot platform URL.

• Enter your username and password.

• Click Log In to access the dashboard.

Step 2: Access the Dashboard

After logging in, the home dashboard appears. Which provides an overview of the platform and quick access to various features.

Step 3: Navigate to Bots

• From the left-side navigation panel, click Bots.

• This opens the Bot Management section where you can create and manage bots.

Step 4: Select the Voice Bot Tab

Within the Bot Management page:

• Click the Voice Bot tab.

• This section allows you to create and manage voice-enabled bots.

Step 5: Enter Bot Name

• Provide a unique and descriptive name for your bot.

• This helps identify the bot later when managing multiple bots.

Step 6: Choose Bot Type

Select the type of voice bot you want to create:

• Inbound Bot – Handles incoming calls from users.

• Outbound Bot – Initiates calls to users for campaigns or notifications.

Step 7: Open the Bot Builder

After creating the bot, the Bot Builder interface will open.

• Start with the Start Component.

• Connect the Start node to the next component in your workflow.

Step 8: Add a Set Variable Component

• Search for the Set Variable component in the component panel.

• Drag it into the flow and connect it to the Start node.

Configure the Variable

Define:

• Variable Name

• Variable Value

This variable can be referenced anywhere within the bot flow.

Step 9: Add the Gen AI Component

• Attach the Gen AI component after the Set Variable component.

• This component enables the bot to generate AI-powered responses based on user input.

Step 10: Add a Prompt (Optional)

While configuring the Gen AI component, you can define a prompt that guides the AI's responses. Currently, adding prompts directly may require:

• Using API tools such as Postman, or

• Assistance from a Product Manager, Developer, or Bot Builder.

Step 11: Select a Prompt from AI Hub

Alternatively, you can select a predefined prompt from the AI Hub. Preconfigured prompts help standardize AI responses across multiple bots. Custom prompt creation can be configured separately within the AI Hub.

Step 12: Configure Query Variable

Set the query variable as: @query.This variable stores user input or spoken queries received during the conversation.

Step 13: Configure Gen AI Response Variable

Set the Gen AI response variable as:

@genai.

This variable stores the AI-generated response. You may also enable filler words to create more natural conversational speech if needed.

Step 14: Configure the Conversation Loop

To maintain an ongoing conversation:

• Connect the output of the Gen AI component to a Custom Input component.

• Connect the output of the Custom Input back to the Gen AI component.

This loop allows the bot to continuously process user queries.

Step 15: Configure Custom Input for TTS Streaming

The Custom Input component is used for Text-to-Speech (TTS) streaming. Streaming AI Responses:

Set the streaming variable as:@genai.

This streams the AI-generated response to the user.

Custom TTS Output:

You can also create additional Custom Input components and manually enter text that should be spoken by the voice bot.

Step 16: Select Language

Navigate to Settings.

• Choose the language your voice bot should use during interactions.

• Language selection determines both speech recognition and response output.

Step 17: Customize TTS Pronunciation

• You can customize Text-to-Speech pronunciation rules for specific words.

• This ensures that brand names, technical terms, or uncommon words are spoken correctly.

Step 18: Configure Speech-to-Text (STT)

• From the side panel, select STT (Speech-to-Text).

• Choose the STT engine that will convert user speech into text.

This allows the bot to understand spoken input during calls.

Step 19: Select the LLM

• Choose the Large Language Model (LLM) that will power your Gen AI responses.

• The LLM processes user queries and generates contextual replies during conversations.

By following these steps, you can create a Gen AI-powered voice bot capable of:

• Understanding spoken user input

• Generating AI-driven responses

• Streaming voice responses through TTS

• Supporting continuous conversational interactions

This setup enables businesses to automate voice interactions and deliver intelligent, real-time voice experiences.