Docs› AI› Voice cloning (Enterprise)

Voice cloning (Enterprise)

Create a custom AI voice for your brand using voice cloning — available on the Enterprise plan.

Available on Enterprise plan only.

Voice cloning lets you replace the default text-to-speech voice with a custom voice that sounds like your brand. Instead of a generic AI voice, your visitors hear a voice you recorded — a voice actor, a brand ambassador, a founder, or anyone whose voice fits your brand identity.

How it works

You provide audio samples. elNudge generates a custom text-to-speech model from those samples. Once approved, that voice is used for all voice-enabled nudges and chat responses on your site.

Steps to set up voice cloning

1. Contact support to enable voice cloning

Voice cloning is enabled by your account manager. Reach out to support or your enterprise contact to activate the feature on your account.

2. Record or source your audio samples

You need at least 10 minutes of clean, consistent audio from a single speaker. More audio (20–30 minutes) produces a higher quality result.

Recording guidelines:

  • Use the same microphone and room for all recordings — inconsistency degrades quality significantly.
  • Record in a quiet environment with no background noise, echo, or air conditioning hum.
  • The speaker should use a natural, steady pace — not too fast, not artificially slow.
  • Vary the content: include questions, statements, product descriptions, and short conversational phrases. Monotone content (reading the same paragraph repeatedly) produces worse results than varied speech.
  • WAV or high-quality MP3 (320kbps) are preferred. Avoid compressed audio formats.

3. Upload via Dashboard → Voice → Custom Voice

Upload your audio files in the Custom Voice section. You can upload multiple files — they are combined to build the voice model. Processing typically takes 24–48 hours.

4. Review and approve the generated voice

Once processing is complete, you will receive an email notification. Listen to the preview samples in the dashboard. If the quality meets your standards, approve the voice. If it does not, you can upload additional audio samples and regenerate.

5. Enable per site

After approval, go to Dashboard → Settings → Voice for each site where you want to use the custom voice. Select your custom voice from the Voice dropdown.

Use cases

  • Brand consistency — a voice your customers associate with your brand, not a generic AI.
  • Local language persona — record a native speaker in Hindi, Tamil, Telugu, or any of the 27 supported languages to create a deeply local experience.
  • Ambassador voice — if you have a brand ambassador, their voice in your AI agent reinforces brand association.

Legal and consent requirements

You are responsible for ensuring you have the right to use any voice you clone. Specifically:

  • Consent — you must have explicit written consent from the person whose voice is being cloned. This applies to employees, contractors, voice actors, and public figures.
  • Usage rights — if the voice belongs to a brand ambassador or celebrity, your contract must grant rights to use their voice in AI-generated speech.
  • Disclosure — in some jurisdictions, you may be required to disclose to visitors that they are interacting with an AI voice. Check applicable regulations for your market.

By uploading audio and generating a custom voice, you confirm that you have the necessary rights and consents. elNudge does not verify third-party rights.