Skip to main content
AudioBook Engine

Privacy

Privacy, manuscripts and voice data

This describes what we collect and what happens to it. It is written to match what the software actually does rather than to cover every eventuality, and it has not yet been reviewed by a lawyer — if something here matters to your decision, email ceo@theappmakerpro.com and you will get a straight answer in writing.

Your manuscript

Most of what you give us is an unpublished book. We treat it that way. The file you upload (EPUB, DOCX or PDF), the chapter text we extract from it, the cover art you attach, and the finished audiobook we render are all stored with public access switched off. None of them is served on the public internet, none is shared with another user, and none appears in any catalogue, feed or search index of ours.

We do not use your manuscript, your cover art, or your finished audio to train any model — not ours, and not a third party's. Your text is sent to a narration service to be spoken and for no other purpose. It is not a training corpus, it is not a sample set, and it is not offered to anyone as either.

The finished files are yours. There is no exclusivity: you can sell the audiobook wherever you like — your own store, Kobo, Spotify, Findaway, Patreon — and nothing about uploading here restricts where it goes.

Manuscript retention

  • Uploaded file and chapter text — kept while the book is in your library, so you can re-render or correct the chapter split without uploading again.
  • Cover art — kept with the book. Replacing it overwrites the previous file.
  • Rendered audio — kept privately so you can download it again. Re-rendering a book replaces it.
  • On request— email us and a book, its text, its cover and its audio are deleted. See “Deleting your data” below for what is and is not self-service today.

Voice recordings

Narrating a book in a clone of your own voice is a separate, optional feature and is not part of rendering an audiobook — you never have to record anything to use this product. If you do clone a voice, this is what happens to the audio.

To clone a voice we ask for at least 60 seconds of speech plus a short spoken consent clip. Both are recorded in your browser and uploaded directly to storage with public access switched off. They are not served on the public internet, they are not used to train any model, and they are never shared with other users of this product.

A cloned voice is private to your account. No other user can narrate with it or hear your samples. The one thing that is public is a short preview clip generated after cloning, so the voice can be auditioned in the picker — it is deleted when you delete the voice.

Retention and destruction schedule

  • Voice samples — kept while the voice exists. Deleted from storage when you delete the voice.
  • Preview clip — public while the voice exists. Deleted when you delete the voice.
  • Consent recording — kept for up to 24 months after the voice is deleted, as evidence that consent was given, then destroyed. This is the term stated in the sentence you read aloud before recording.
  • Abandoned uploads — if you start the wizard and do not finish, or a sample is rejected, the uploaded audio is destroyed within 24 hours.
  • Consent timestamp and IP address — recorded with the consent clip and kept for the same period, for the same reason.

Only clone a voice that is yours, or one you have the narrator's permission to use.

Who else processes your data

  • Modal — runs the narration model that speaks your chapter text
  • Clerk — accounts and sign-in
  • Stripe — payments (we never see your card number)
  • Neon — the database
  • Cloudflare R2 — manuscript, cover and audio storage
  • ElevenLabs — only if you clone a voice. It receives your voice samples and holds the clone. It never receives your manuscript.

Deleting your data

You can delete a cloned voice yourself at any time, from the Voices page. There is not yet a self-service way to delete a book, delete your whole account, or export your data — that is a gap we are naming rather than papering over. Email ceo@theappmakerpro.com and it will be done by hand. If you ask us to delete everything, the consent recordings are the one thing we may keep to the end of the period above, because they are the proof that a real person agreed to their voice being cloned. A manuscript has no such exception: ask, and it goes.

Our terms of service are still being drafted. Pricing, refunds and what happens to a failed render are described on the pricing page and are accurate today.