Last updated · June 13, 2026

Privacy Policy

30K is a personal diary. Your entries belong to you — we built the app around that principle. This page explains, in plain language, what we collect, why, and who else touches it.

30K is operated by Andrei Blokhin, an individual based in Cyprus (“we”, “us”). This policy covers the 30K iOS app and related services. By using 30K you agree to what’s described here.

What we collect

Account. When you sign in we store an account identifier. If you use Sign in with Apple, we receive the relay email Apple gives us (we don’t see your real address unless you share it). You can also use 30K anonymously — in that case we generate a random ID with no personal info attached.

Your entries. The diary content you write — text, dates, voice notes — is stored in your account so it can sync across your devices. We do not read it manually. It is used only to power the app’s features for you.

Opt-in AI features. The AI features below — voice transcription, voice tone/context, photo recognition, and entry analysis (reflections, chapters, outlines, search) — are turned on or off by you. During onboarding you’re asked explicitly, and you can change your choices at any time in Settings. We send only what a given feature needs to work, and only to the provider that powers it (each is named in the Third parties section below). If a feature is off, nothing is sent to any AI provider for it. And whatever is sent carries the content alone — the entry text, the audio, or the photo — never tagged with your name, email, or account. A provider sees what a feature needs to process, not who you are.

Voice notes. If voice transcription is enabled and you record audio, the audio file is sent to two providers: Deepgram, which converts your speech to text, and Google’s Gemini API, which analyzes tone and surroundings (mood, non-speech sounds) to add context to the entry. The audio is then stored alongside your entry. If transcription is disabled, the audio is still saved with your entry, but it isn’t sent to either provider.

Photos. If photo recognition is enabled and you attach photos to an entry, the image files are uploaded to your account and sent to Google’s Gemini API so the app can read text in the image, describe what’s on it, and tie it into your entry. We also read the photo’s metadata (EXIF) — image size, capture date, device, and GPS coordinates if the photo file contains them — and use it as context for the same purpose. Photos and their metadata are stored alongside the entry. If photo recognition is off, photos are still stored with your entry but aren’t sent to Google.

Reflections, chapters, and search. If you enable the assistant and book features, the text of your entries — together with any voice transcripts and photo descriptions tied to them — is sent to Google’s Gemini API to generate reflective questions, chapter drafts, and outlines, and to OpenAI to build a private search index (an “embedding” — a numeric representation that powers search and surfaces related past entries). If these features are off, this text is sent to neither.

Audiobook narration. When you generate narration for a chapter, or hear a question read aloud in the voice interview, that text is sent to Google’s Gemini API to be voiced (text-to-speech) in one of a few preset narrator voices you choose; the resulting audio is stored in your account so it doesn’t have to be remade. This uses a preset voice — nothing about your own voice is cloned. Any answer you record in the interview is handled as a voice note, exactly as described above.

Importing a diary. If you import an existing diary as a file and we can’t read its dates automatically, the relevant date lines may be sent to Google’s Gemini API to interpret them. The rest of the import runs on our own servers.

Device & usage data. App version, OS version, device model, language, crash and error reports, and product analytics events (which screens you open, which buttons you tap). We use this to fix bugs and improve the app. No diary content is included in crash reports or analytics events.

Purchases. Subscription status comes from Apple via StoreKit. We don’t receive your card details — Apple handles payment.

Push tokens. If you enable notifications, we store an APNs token to deliver them.

What we don’t collect

No contacts, no health data, no advertising IDs. We don’t track your device’s location — the only location data we ever see is the GPS coordinates baked into a photo’s EXIF, if you choose to attach a photo that has them. 30K asks for two iOS permissions: the microphone, only when you record a voice note, and the camera, only when you take a photo or selfie to add to an entry. Existing photos are added through Apple’s system picker (or the iOS share sheet), which needs no photo-library permission and only hands us the files you explicitly select.

How we use your data

  • Run the app’s core features (writing, syncing, search, reminders).
  • Power AI features you’ve enabled — reflections, chapter generation, outlines.
  • Transcribe voice notes you record, if voice transcription is enabled.
  • Read and describe photos you attach so they become part of the entry, if photo recognition is enabled.
  • Build a search index over your entries so you can find related ones, if book features are enabled.
  • Read your chapters and interview questions aloud (text-to-speech), if you use the audiobook or voice interview.
  • Send you push notifications you’ve opted into.
  • Diagnose crashes and understand which features are used.
  • Process subscriptions and prevent abuse.

We do not sell your data, and we do not use the content of your entries for advertising or to train AI models — neither our own models nor our providers’.

Third parties

We keep this list short on purpose. Every provider below is bound by a data-processing agreement, receives only the minimum a feature needs, and is contractually barred from training on your content or keeping it longer than its task requires.

  • Apple — Sign in with Apple, In-App Purchases, push notifications (APNs).
  • Google (Gemini API) — used only for AI features you’ve enabled: voice tone/context, photo recognition, entry analysis (reflections, chapters, outlines), and reading your chapters and interview questions aloud (text-to-speech). It also interprets dates when you import a diary file. The relevant audio, image, or entry text is sent to produce a response. If a feature is off, nothing goes to Google for it. We use the Gemini API under its no-training policy for API traffic; Google may retain data briefly for abuse monitoring per their terms.
  • Deepgram — speech-to-text. When voice transcription is enabled, the audio of a voice note is sent to Deepgram and a transcript is returned. Deepgram does not use your audio to train its models.
  • OpenAI — used only when book features are enabled: the text of your entries is sent to OpenAI to build a private search index (embeddings). OpenAI excludes API data from model training.
  • Sentry — crash and error reporting. Receives crash reports, error events, and your account identifier so we can match a crash to the affected user when investigating. No diary content is sent.
  • PostHog — product analytics. Receives device/usage events as described above. No diary content is sent.

We may add or change providers as the app evolves; this list is kept current, and significant changes are announced in-app.

Our backend (database, sync, auth) runs on infrastructure we control — Supabase and PowerSync, both self-hosted on our own servers. Your data is not shared with Supabase Inc. or PowerSync Inc.

Security

Data is encrypted in transit (TLS) between your device and our servers. Our servers run on encrypted disks and access is limited to the developer operating the service. We do our best, but no system is perfectly secure — if you spot something, please email us.

Data retention and deletion

You can delete your account at any time from within the app (Settings → Delete account). Deletion is immediate and permanent — your entries, audio, and account record are removed from our database. Backups roll off within 30 days.

You can also export everything as a ZIP archive from the app at any time, and import an archive back later.

If you stop using 30K without deleting the account, we keep the data so you can come back to it.

Your rights

You can:

  • See your data — it’s all in the app, or exportable as a ZIP.
  • Correct or delete any entry, individually or all at once.
  • Delete your account and all associated data.
  • Ask us a question about your data — email below.

If you live in the EU/UK, you have rights under GDPR (access, rectification, erasure, portability, objection). The above features cover all of them; if you’d rather we handle it manually, email us.

Children

30K is intended for users 13 and older — or the minimum age required by your local law (for example, 16 in parts of the EU/UK). We don’t knowingly collect data from anyone younger. If you believe a child has created an account, email us and we’ll delete it.

Changes to this policy

We may update this policy as the app changes. The “updated” date at the top reflects the latest version. Significant changes will be announced in-app.

Contact

Questions, requests, or anything about your data — email [email protected].