Privacy Policy
Last updated: February 7, 2026
Summary
Dispatch records voice and saves it to your device. If you choose to connect Google Drive, recordings are uploaded to your own Drive account. No analytics, no tracking, no ads, no accounts with us.
1. Data Collection
Dispatch does not collect, transmit, or store any personal data on our servers. We have no servers. We never see your recordings.
2. Recordings & Transcription
All voice recordings are saved locally on your device in a folder you choose (default: Documents/dispatch/). Recordings are stored as .m4a files. Dispatch does not access, read, or modify any files outside of this folder.
If on-device transcription is enabled, a language model (Moonshine Base) is downloaded to your phone and all transcription happens locally. The model download is a one-time network request to HuggingFace; after that, transcription requires no internet connection.
3. Google Drive (Optional)
You can connect your Google account in settings to upload recordings to your own Google Drive. This is entirely optional.
- Dispatch uses the
drive.filescope — it can only access files that Dispatch itself created. It cannot see, read, or modify any other files on your Drive. - Recordings are uploaded to a
dispatch/folder on your Drive. - Your Google account email is stored locally on your device to maintain the connection. It is not sent to us or anyone else.
- You can disconnect Google Drive at any time in settings. Dispatch reverts to local-only mode.
4. Permissions
Dispatch requires:
- Microphone — required to record audio.
- Notifications — required by Android to keep recording when the screen is off.
- Storage — required on older Android versions to save recordings to the Documents folder.
- Internet — used when Google Drive is enabled and for the one-time download of the transcription model (if on-device transcription is enabled). No ongoing network requests are made after the model is downloaded.
5. Third-Party Services
When Google Drive is enabled, Dispatch uses Google's sign-in and Drive APIs to upload recordings to your account. When on-device transcription is enabled, the Moonshine model is downloaded from HuggingFace (one time). No other third-party SDKs, crash reporting, ad networks, or analytics are used.
6. Data Sharing
Dispatch shares no data with us or any third party. When Drive is enabled, recordings go to your Google account — that's between you and Google.
7. Contact
Questions about this policy: support@newyorkai.org