Privacy Policy
Effective date: 2026-09-14
Voice Bio ("we", "our", "the app") is a voice journaling app that helps you record spoken thoughts, turn them into organized text, and weave them into storylines and autobiographies. This policy explains what we collect, why, who processes it on our behalf, and the choices you have. It applies to the Voice Bio iOS, Android, web, and desktop apps.
If you have questions, email us at jackychung@voice-bio.com.
1. Who is the data controller
Voice Bio is operated by Jacky Chung, with correspondence at jackychung@voice-bio.com. You can reach the data controller at jackychung@voice-bio.com.
2. What we collect
2.1 Account data
When you sign in, we store:
- Your email address (required for email magic-link sign-in)
- Your provider user ID if you use Sign in with Apple or Sign in with Google (we do not see your password with these providers)
- If you use Sign in with Apple, a refresh token issued by Apple, stored only so we can revoke it when you delete your account
- A display name you choose (optional, editable in Settings)
- Your interface language and preferred output language
- Your birth year (optional, used only to frame autobiography prompts)
2.2 Content you create
- Voice recordings you make in the app
- Transcripts generated from those recordings
- Organized / polished text produced by AI from your transcripts
- Storylines, chapters, autobiographies, and quotes you create or that we help generate from your entries
- Tags, moods, and titles you attach to entries
All of the above is scoped to your account. Other users cannot see it. We do not use your content to train AI models.
2.3 Purchase data
If you subscribe, our payments partner RevenueCat receives:
- A pseudonymous app user ID (your Voice Bio user ID)
- The product you purchased and the receipt issued by Apple or Google
- A country code inferred from your store account
We do not receive your payment card or Apple ID / Google account details — those stay with Apple and Google.
We store account-linked monthly transcription seconds, AI request counts, estimated processing costs, and request status for allowance management, duplicate prevention, and cost analysis. Deleting a journal entry does not reset monthly usage. RevenueCat provides subscription revenue analytics and entitlement validation; it does not receive journal text or recordings.
Photos you choose for an entry are resized on your device and stored with your account in the private Supabase entry-photos bucket. We store image dimensions, file size, and display order. Photos are displayed using temporary signed links and are removed when you delete the photo, its entry, or your account. This feature does not send photos to AI providers.
Edit history stores earlier text versions of your entries and quotes, including titles, tags and AI writing results, in your account. Only you can read these versions. Deleting an entry, quote or account also deletes its corresponding history. Photo and audio files are not copied into text history.
2.4 Device permissions
- Microphone — used only while you are actively recording. The app does not record in the background and does not access the microphone when you are not on the recording screen.
3. What we do NOT collect
- We use RevenueCat for subscription revenue analytics and validation. We do not use tracking pixels or advertising SDKs.
- We do not track you across other apps or websites.
- We do not sell your data.
- We do not use your voice or text content to train our own or any third party's AI models.
4. Third-party processors
We rely on the following services to run Voice Bio. Each of them processes data only on our instructions and subject to their own terms.
| Processor | Purpose | What they receive |
|---|---|---|
| Supabase | Hosting, authentication, database, file storage, Edge Functions | Account data, content, audio files, Apple refresh token |
| Resend | Delivering account sign-in and authentication emails | Recipient email address, authentication email including its sign-in link, delivery status |
| OpenAI | Audio transcription (gpt-4o-transcribe, whisper-1) and text generation (gpt-5-mini) | The audio file and/or text for that single request |
| Apple | Sign in with Apple; App Store billing | Identity assertion; purchase receipts stay with Apple |
| Sign in with Google; Play billing (Android) | Identity assertion; purchase receipts stay with Google | |
| RevenueCat | Subscription receipt validation and entitlement management | App user ID, store receipt, product identifier, store country |
| Vercel | Public marketing and legal website only | No account or journal data |
Resend delivers authentication emails through our Tokyo sending region and processes email delivery records. We do not send journal content or audio to Resend, or enable email open/click tracking for sign-in emails.
OpenAI states that API inputs are not used to train their models under default paid API terms. Consult their policies for the authoritative statement. We do not opt your content into any training programme.
4.1 Supabase
- Auth stores your email or Apple/Google provider identity and issues session tokens. Passwords for Apple/Google sign-in never reach us.
- Postgres stores your profile, entries, transcripts, organized text, storylines, chapters, quotes, autobiographies, and (if you used Sign in with Apple) the Apple refresh token used only for account-deletion revocation.
- Storage holds voice files in a private
audiobucket ataudio/{userId}/{entryId}.m4a. Row-level security and storage policies restrict access to your own folder. - Edge Functions read only the row needed for the request, call OpenAI / RevenueCat / Apple as described here, and write results back to your rows.
- Region: AWS Asia Pacific (Sydney),
ap-southeast-2. - Client access is limited by row-level security. The service role is used only inside Edge Functions after verifying your JWT.
4.2 OpenAI
- Transcription:
gpt-4o-transcribereceives the audio file for that entry. If you marked quote moments while recording, we also send the same file towhisper-1for word-level timestamps. - Text:
gpt-5-minimay receive transcript or journal text to organize an entry, polish a transcript or quote, suggest storylines, compile or regenerate autobiography chapters, or pick a shareable excerpt. - OpenAI receives only the payload for that request — not your full library, password, or payment details. We do not grant OpenAI a right to train on your content.
4.3 RevenueCat
- The iOS/Android SDK logs in with your Voice Bio (Supabase) user ID as the RevenueCat
app_user_id. - RevenueCat receives that ID, the App Store or Play Store receipt, the product identifier (
vb_premium_monthly,vb_premium_yearly, orvb_lifetime), entitlement status, and a country/locale inferred from the store account. - Edge Functions call RevenueCat's REST API to enforce the monthly transcription, text-processing, and book-generation allowances and to cache
planon your profile (service role only). - RevenueCat does not receive recordings, transcripts, journal text, or your card number.
- When you delete your account we call RevenueCat's subscriber delete API. This does not cancel the store subscription.
5. Where your data is stored
Your content is stored in Supabase in AWS Asia Pacific (Sydney), ap-southeast-2. Audio files are held in a private Supabase Storage bucket; only you (and our backend acting on your behalf) can read them.
OpenAI processes a request in their own infrastructure and does not become the system of record. RevenueCat stores entitlement records in its own infrastructure. The public site at voice-bio.com is static HTML on Vercel and does not store journal data.
6. How long we keep it
- While your account exists, we keep your content so you can use and re-read it.
- When you delete your account (see Section 8), we immediately delete the account record, profile, entries, transcripts, organized text, storylines, chapters, quotes, autobiographies, and audio files; revoke the Apple Sign in token if one was stored; and delete the RevenueCat subscriber record.
- Backups may retain residual copies for up to 30 days before being overwritten.
7. Legal bases (for EEA / UK users)
- Performance of a contract — to give you the app you signed up for (storing your entries, generating transcripts, etc.).
- Legitimate interests — to keep the service secure and prevent abuse.
- Consent — for microphone access, which you grant at the OS level and can revoke at any time in system settings.
8. Your rights and choices
You have the right to:
- Access — view all of your data from within the app.
- Correction — edit your display name, entries, and other content in-app.
- Deletion — delete your account from Settings → Delete account. This is permanent and removes all server-side data, including audio files, and revokes your Sign in with Apple token when one exists.
- Portability — request an export by emailing jackychung@voice-bio.com.
- Withdraw consent — revoke microphone permission in your device settings.
- Complaint — if you are in the EEA or UK, you may lodge a complaint with your local data protection authority.
Deleting your Voice Bio account does not cancel your Apple App Store or Google Play subscription. Cancel the subscription separately at Settings → Apple ID → Subscriptions (iOS) or the Play Store (Android).
9. Children
Voice Bio is not intended for children under 13 (or under 16 in the EEA). We do not knowingly collect data from children. If you believe a child has given us data, email jackychung@voice-bio.com and we will delete it.
10. Security
We use Supabase's row-level security to ensure your data is only accessible to you. Audio files are held in a private bucket. All traffic is encrypted in transit (HTTPS / TLS). No system is perfect — if you suspect a breach of your account, contact us.
11. Changes to this policy
If we change this policy in a material way, we will notify you in the app or by email before the change takes effect. The "Effective date" at the top will always reflect the current version.
12. Contact
jackychung@voice-bio.com