#

How to add or modify voice of AI Avatar

Created by Ayush Sharma, Modified on Tue, 30 Sep at 3:24 PM by Ayush Sharma

Your AI avatar is more than just a visual representation it’s the voice that communicates with learners or users. The way your avatar sounds can dramatically influence engagement, clarity, and trust. Choosing the right voice helps you:

  • Personalize communication – A voice similar to yours or your brand tone makes interactions feel authentic.

  • Increase engagement – Natural and pleasant voices keep learners attentive and encourage longer interaction.

  • Build trust – A consistent voice reinforces reliability and makes the AI feel human-like.

  • Set the right mood or style – Calm, energetic, confident, or storytelling voices can influence how users perceive the learning experience.

This guide explains how to choose a pre-made voice or train your avatar to speak like you, step by step, with reasoning behind each action.

Step 1: Navigate to Products → AI Avatar → View Details.


Step 2: In the Profile section, click Manage Voice.

Step 3: Choose Your Voice Option

  • Recommended: Your Own Voice – Lets your avatar speak just like you.
    Personalization helps learners feel a stronger connection to the avatar.

  • Other Options: Pre-Made Voices – Choose from voices like Christina, Jason, Raghav, or Karishma.
    These voices are pre-trained for clarity, style, and emotion. They’re ideal if you want a specific tone or don’t want to record your own voice.

Step 4: Add Your Own Voice (if selected)

  • Click + to start adding your own voice.

  • You can either Upload files or Record audio.


Upload Voice Files:

  • Drag & drop audio files or click to upload (max 10 MB each). Uploading existing high-quality recordings can save time and provide a clear sample for training.

Record Audio:

  • Read aloud from the provided sample content. Live recording ensures your avatar learns your natural pronunciation, tone, and speaking style.

Recording Instructions:

  1. Quiet environment: Avoids background noise, which could make the AI misinterpret sounds.

  2. Length: 30 seconds to 2 minutes provides enough data for accurate voice training.

  3. Use your real voice: Playing pre-recorded audio doesn’t capture natural nuances, leading to a robotic or unnatural avatar voice.

Sample Audio:

  • You can play Sample Recording.mp3 as a reference for style and pacing. Following the sample ensures consistency and helps your avatar replicate a clear, natural tone.

Step 5: Review and Save

  • After uploading or recording, your audio appears under Your Audio Samples.

  • Review to make sure the quality and clarity are good.

  • Clean Up Your Audio

    After recording/uploading, you’ll also see an option to remove background noise - Enable this to automatically filter out ambient sounds like fans, typing, or street noise—ensuring your avatar's voice sounds clear and professional.

    Use this only if the audio contains noticeable noise.


  • Click Save 

  • This final step confirms that the avatar will now use the trained voice in conversations. You can always retrain or update later as your style or content changes.


Step 6 : Configure Voice Settings

After adding a voice, you can fine-tune how your avatar speaks:

  1. Language

    • Choose the language in which your avatar will speak (e.g., ?? English). Ensures pronunciation and expressions match the language of your learners.

  2. Call Greetings / Initial Message

    • A default greeting is already set, e.g., “Hello! How can I assist you today?”
    • You can edit it if you want a more personalized or brand-specific opening.
    •  A friendly, clear greeting makes learners feel welcomed and sets the tone for the conversation.
  3. Fallback Message on Call End

    • A default fallback message is already set, e.g., “Looks like you’ve stepped away. Please call me again if you want to chat.”
    • You can edit it if you want a more personal or context-specific message.
    •  This message maintains engagement and politely informs learners when the session is ending, especially if they are silent.
  4. Speed
    Control how fast your avatar speaks using the adjustable slider.

    • Move slider to the right (Faster)
       → The avatar will speak more quickly.

    • Move slider to the left (Slower)
       → The avatar will speak more slowly. 

  5. Stability
    Voice Stability Adjust how expressive or flat the voice sounds during conversations.

    • Move slider to the right (More stable)
      → Voice becomes smoother and consistent.

    • Move slider to the left (More variable)
      → Voice has more natural fluctuations and emotion. 

  6. Style Exaggeration
    Style Exaggeration Enhance voice dramatics for storytelling or emphasis when needed.

    • Move slider to the right (More exaggerated)
       → Adds more pitch variation, emotion, and excitement. 

    • Move slider to the left (Less exaggerated)
       → Voice remains neutral and professional. 

  7. Reset All Values

    • Restore default settings if changes don’t give desired results. Provides a safe way to start over and experiment without losing all configurations.

Step 7: Test Before Enabling

  • Before learners can hear your avatar, voice will be disabled by default.

  • Test your voice to ensure it sounds right.
    This ensures that learners experience clear, natural speech and gives you a chance to fix any issues before enabling it publicly.


Was this article helpful?

That’s Great!

Thank you for your feedback

Sorry! We couldn't be helpful

Thank you for your feedback

Let us know how can we improve this article!

Select at least one of the reasons
CAPTCHA verification is required.

Feedback sent

We appreciate your effort and will try to fix the article