Skip to main content

Cloned Voices

Written by Alyssa Elso

What are Cloned Voices?

Voice Cloning lets your practice create a custom AI voice from audio samples of a real person's speech, such as a doctor or office manager. You can then use that voice for voicemail and other patient communications, so messages sound like someone patients already know.

If you don't want to create a custom voice, PbN also offers built-in system voices. These are professional voices that are ready to use right away.

How voice cloning works:

  1. Voice sample collection. You provide 2–3 minutes of clear, high-quality speech. Several samples give the best accuracy.

  2. AI voice learning. PbN analyzes the samples, captures the speaker's unique voice characteristics and speech patterns, and trains a voice profile.

Where to find Cloned Voices

Go to Settings > Voice, then select Cloned Voices under Voice Management. You can also type "Cloned Voices" into the settings search bar.

The Voice Cloning page includes:

  • How It Works opens an overview of how voice cloning works, plus privacy details and best practices.

  • Refresh reloads the page so you can see the latest statuses.

  • Create Voice Clone starts a new custom voice.

  • Filter by type shows Custom Voices, System Voices, or both.

  • Filter by status shows voices that are Pending, Training, Ready, or Failed.

  • Custom Cloned Voices shows the personalized voices your practice has created from audio samples. Each card shows the voice name, its status, how many audio files it has, and Preview Voice, Edit, and Delete buttons.

  • Built-in System Voices shows PbN's pre-configured professional voices. They come in a range of styles, such as professional, narrator, warm, energetic, deep, or gentle. Click Preview Voice on any card to hear it.

Practice by Numbers Voice Cloning settings page showing type and status filters and a custom cloned voice card

Built-in System Voices in Practice by Numbers voice cloning with Preview Voice options

Voice statuses

Status

What it means

Pending

The voice has been created and is waiting for audio or processing.

Training

PbN is learning the voice from your samples.

Ready

The voice is ready to use. On the card this shows as Ready To Use.

Failed

The voice couldn't be created. Try again with clearer, higher-quality audio.

How to create a cloned voice

Before you start, get permission from the person whose voice you're cloning, and record their audio samples (see Best practices below).

  1. Click Create Voice Clone.

  2. In the Create New Voice window, enter a Name for the voice. Use a name your team will recognize, such as the person's name or role.

  3. Optional: add a Description.

  4. Click Create Voice.

  5. Add the person's audio samples. Aim for 2–3 minutes of clear speech in total.

Create New Voice window for creating a cloned voice in Practice by Numbers with Name and Description fields

The new voice appears under Custom Cloned Voices. Its card shows how many audio files have been uploaded. The status moves from Pending to Training while PbN learns the voice, then to Ready To Use.

How to preview a voice

Click Preview Voice on any custom or system voice card to hear a sample before you use it.

How to edit a cloned voice

  1. Click Edit on the voice's card.

  2. In the Edit Voice window, update the Name or Description.

  3. Click Update Voice.

Edit Voice window for renaming a cloned voice in Practice by Numbers

How to delete a cloned voice

Click Delete on the voice's card and confirm. Delete a cloned voice when you no longer need it, for example when the person leaves the practice or asks you to remove it.

Built-in System Voices can't be edited or deleted.

Best practices for recording samples

  • Record in a quiet room. If you can, use a good-quality microphone. Background noise lowers the quality of the cloned voice.

  • Speak naturally. Talk clearly, at a normal pace and tone, the way you'd speak to a patient.

  • Vary what you say. Use a mix of phrases and expressions rather than repeating the same sentence.

  • Provide enough audio. 2–3 minutes of clear speech, split across several samples, gives the most accurate results.

Privacy and security

  • Encrypted storage: voice data is encrypted at rest.

  • Access control: strict permission controls limit who can manage voices.

  • Secure transfer: voice data is protected with end-to-end encryption.

  • Data removal: you can delete a cloned voice at any time.

Frequently asked questions

  1. What's the difference between a custom voice and a system voice?
    A custom voice is cloned from audio of someone at your practice. A system voice is a pre-configured professional voice that PbN provides, and you can use it right away without recording anything.

  2. My cloned voice failed. What should I do?
    This is usually caused by audio that's too short, noisy or unclear. Delete the voice and create it again with 2–3 minutes of clean speech recorded in a quiet space.

  3. How long does training take?
    The card shows Training while PbN learns the voice. Click Refresh to check whether it has changed to Ready To Use.

  4. Can I rename a cloned voice?
    Yes. Click Edit, change the name and click Update Voice.

Did this answer your question?