Back to home

Privacy Policy

Last updated: May 11, 2026

Hone is an iOS app that turns voice brain-dumps into structured business ideas. This policy describes what data Hone collects, what it does with that data, and what choices you have. The app is built and operated by Jakub Pilch.

If you have questions about anything in this policy, email jakub.pilch.01@gmail.com.

TL;DR

  • Your audio never leaves your iPhone. Recording and speech-to-text both run on-device.
  • Transcript text only leaves your device if you opt into a cloud AI model — in which case it's sent to Anthropic (via OpenRouter) for summarization and chat replies, and is never stored on our servers or used to train AI models.
  • We don't have user accounts. We don't sell or share your data. We don't use third-party advertising, tracking pixels, or analytics SDKs beyond the two listed below.
  • Crash reports (Sentry) and anonymous product analytics (TelemetryDeck) are collected to help us fix bugs and understand which features people use. Both are GDPR-compliant and store no personally identifying data.

What we collect, when, and why

1. Recordings and transcripts (stays on your device)

When you record an idea, the audio is captured locally and transcribed to text using Apple's on-device Speech framework. The audio file is processed in memory and discarded; only the resulting text transcript is saved to your iPhone's local database.

The recorded audio never leaves your iPhone. It is not uploaded to Hone, to Apple, or to any third party.

The text transcript is stored locally on your device and synced via iCloud (your personal iCloud account) if you have iCloud sync enabled for the app. Apple's iCloud is governed by Apple's own privacy policy.

2. Cloud AI processing (opt-in, transcript text only)

If you choose to use a cloud AI model — either explicitly in Settings or because Apple's on-device model doesn't support your spoken language — Hone sends the text transcript of your recording to our backend, which forwards it to Anthropic's Claude model via OpenRouter for analysis.

We ask for your explicit consent the first time this happens, in the "Cloud AI processing" sheet inside the app. Until you accept that sheet, no text leaves your device.

What happens to the text on our backend:

  • It's processed by Anthropic's Claude model and a structured response is returned to your iPhone.
  • We log only token counts and cost to our database (Supabase), attached to an anonymous Apple transaction identifier. We do not store the transcript text itself, the model's response, or any information that would let us reconstruct what you said.
  • Anthropic's stated policy is that API requests are not retained beyond the duration of the request and are not used to train AI models. See Anthropic's privacy policy at anthropic.com/privacy for their commitment.
  • OpenRouter acts as a passthrough router; their policy is at openrouter.ai/privacy.

3. Subscription and purchase data (from Apple)

When you purchase a Pro subscription, Apple's StoreKit issues a signed transaction receipt. Our backend verifies that receipt against Apple's public certificate chain to confirm you have a valid subscription. We store only the anonymous Apple transaction identifier — never your Apple ID, email, name, billing address, or payment details. Those remain entirely with Apple.

Apple also notifies our backend when key subscription events happen (initial purchase, renewal, cancellation, refund). We use these events to keep your Pro entitlement in sync.

Apple's handling of your purchase is governed by Apple's own privacy policy at apple.com/legal/privacy.

4. Crash reports and diagnostics

We use Sentry to collect crash reports and app-hang diagnostics from released versions of the app (TestFlight and App Store builds). When a crash happens, Sentry collects:

  • The stack trace of the crash
  • iOS version, device model, free memory at the time of the crash
  • A randomized device identifier generated by Sentry — this is not your Apple ID, IDFA, or any cross-app identifier

Sentry does not collect your transcripts, recordings, or any content you typed. Sentry's privacy policy is at sentry.io/privacy.

5. Anonymous product analytics

We use TelemetryDeck to collect anonymous product signals so we can see which features are used and where users get stuck. TelemetryDeck is based in Germany and was built specifically to be GDPR-compliant without consent banners.

Signals we send include events like:

  • app.launched
  • recording.started, recording.completed
  • summary.generated, summary.failed
  • chat.message_sent, chat.response_received
  • paywall.shown, paywall.purchased

Each signal carries the event name and minimal metadata (e.g. word count, duration in seconds, error reason category). It does not carry your transcripts, recordings, chat messages, or any identifying information.

TelemetryDeck creates an anonymous user hash from your iOS vendor identifier, hashed locally on your device so the raw identifier never reaches their servers. Their privacy policy is at telemetrydeck.com/privacy.

6. Feedback and roadmap

If you submit feedback or vote on the public roadmap (linked from Settings), that feature is hosted by Featurebase. Their privacy policy applies to anything you post there. You may optionally provide your email — that's the only personal information they receive, and only if you choose to enter it.

What we don't do

  • We don't have user accounts and don't ask you to register.
  • We don't sell your data.
  • We don't share your data with advertisers, data brokers, or any third party for marketing purposes.
  • We don't use third-party advertising SDKs.
  • We don't use cross-app tracking, fingerprinting, or IDFA.
  • We don't use cookies (there's no Hone website you log into).

Your rights

Because we don't store the content of your transcripts or any personally identifying information, there's nothing user-specific to delete from our servers. The data lives locally on your iPhone — you can delete it any time by deleting a jot inside the app or by uninstalling the app.

You have the right under GDPR (if you're in the EU) and similar laws elsewhere to ask what data we hold about you and to request its deletion. Email jakub.pilch.01@gmail.com with such a request.

Children

Hone is not directed at children under 13. We don't knowingly collect personal information from children under 13. If you believe a child has provided us with personal information, please email jakub.pilch.01@gmail.com and we will delete it.

International transfers

Our backend is hosted on Supabase's EU servers (eu-central-1, Frankfurt). Sentry and Anthropic process data on their own infrastructure, which may include the United States. By using Hone's cloud features, you consent to your transcript text being processed in those regions for the duration of the API request.

Changes to this policy

If we make material changes to this policy, we'll update the "Last updated" date at the top and (for significant changes) notify users inside the app. Continuing to use Hone after a change means you accept the updated policy.

Contact

Jakub Pilch
jakub.pilch.01@gmail.com