Skip to content
Stille Post
Appearance
Colors
Color scheme

Privacy policy

Last updated: November 1, 2026

The short version

Stille Post is an app that supports deafblind people and people with combined hearing and sight loss in communicating. It transcribes speech to text, recognizes different speakers and can translate the text into other languages.

The app normally processes audio and text only on your device. There are no ads and no analytics.

Data is sent over the internet in three cases only: when you download speech or translation models, when you turn on transcription through a server, and when you contact us from the app. For this we also use service providers. The details are below.

Who is responsible?

The controller under the EU General Data Protection Regulation (GDPR) is:

Deutsche Gesellschaft für Taubblindheit gGmbH
Hollestraße 1, 45127 Essen, Germany
Phone: +49 201 26677047
Email: info@gesellschaft-taubblindheit.de

Data protection officer:

DataGAP GmbH
Bessemerstraße 51, 1st floor, 12103 Berlin, Germany
Email: datenschutz@gesellschaft-taubblindheit.de

On your device

  • Audio and text: The app turns speech into text and translates it if you want. Audio, text and translation stay in memory. You can save a transcript via “Show full transcript > Save transcript”.
  • Speakers and messages: The app can tell voices apart during a recording. If you choose “Remember voice” for a speaker, the app stores a voice profile with the name you assigned. A voice profile is a series of numbers. It contains no sound. With this profile, the app can recognize the person by their voice later. Voice profiles are biometric data. They are specially protected under Art. 9 GDPR. So only store a voice profile if the person has explicitly agreed. The app asks you about this first. The voice profile is not sent to us. We have no access to it. It is stored encrypted in the app's storage and is not in the diagnostic files. Whether it is included in backups depends on your device. This is described below under “Encryption and backups”. You can delete it at any time under “Settings > Speaker recognition”. A voice on your device reads typed messages aloud. If you wish, the app translates the message on the device first. Nothing of the messages is stored.
  • Settings: Your settings, the device identifier and the credentials for the server live in the app's storage. The device identifier with the device key and the credentials are encrypted.
  • Diagnostic files: The app records technical values, such as levels, timings and error messages. No audio and no transcript text is in them. The files leave the device only as an attachment to a contact email, if you want that.

If you delete the app, the data in the app's storage is deleted. Transcripts you saved outside the app remain. You have to delete them yourself.

Encryption and backups

The app stores three kinds of data encrypted: voice profiles, the device identifier with the device key, and your credentials for the server. For this the app uses AES-256-GCM, a recognized method. The protected key store of your operating system keeps the key. Anyone who only copies the app's files cannot read them without this key.

The device identifier, the device key and the credentials belong to exactly one device. That is why they are not included in backups on iPhone, iPad, Mac and Android. For voice profiles, the following applies depending on the device:

  • iPhone, iPad and Mac: Apple's keychain keeps the key. Voice profiles are included in backups, for example in iCloud, on your computer or in Time Machine. They are encrypted there too.
  • Android: The Android keystore keeps the key. This key never leaves the device. That is why voice profiles are not included in backups.
  • Windows and Linux: On Windows, your user account protects the key. On Linux, the keyring of your desktop environment keeps it. The app does not exclude any files from backups here. If you back up your device, all three kinds of data are included in the backup, encrypted.

Some devices offer no such key store. This affects the Mac version that does not come from the App Store, and Linux without a keyring. The app then puts the key in a file in the app's storage. Only your user account can read this file. On the Mac, the file is not included in backups. This protection is weaker: anyone with access to your user account can read the data.

If you restore a backup on a device where the key is missing, the app cannot open the voice profiles. It then deletes them.

On iPhone, iPad and Mac, the key can remain in the keychain after you delete the app. It is a random number and contains none of your data.

Downloading the speech models

The app needs speech models to transcribe. On the Mac the speech and speaker models are part of the app. On iPhone, iPad, Android, Windows and Linux the app downloads them from our server models.almost-senseless.tech once you start that. Stille Post downloads the translation model from the same server on every device once you turn translation on. The app downloads updated models from our server on every device. That too happens only when you explicitly start it.

The server receives your IP address in the process, as with any download. The server does not log the downloads. Only the server's error messages can contain the IP address. We delete them after 14 days. The legal basis is our legitimate interest in providing the app with current speech models (Art. 6(1)(f) GDPR).

Transcription through a server

By itself the app works without the internet. Audio is transmitted only when you switch that on. The switch is called “Through a server”. You find it under “Settings > Speech to Text”.

Our server stille-post-server.almost-senseless.tech is preset. While recording, the app sends the audio there. It also transmits the chosen language, the device identifier and a digest of the device key. If you entered personal credentials, it transmits your user name and a digest of your password instead. With translation on, it also sends the target language. The server also receives your IP address. It returns the recognized text and, if requested, the translation, and stores neither audio nor text. We process the audio, the text and the translation on your behalf (Art. 28 GDPR). We do not use them for our own purposes. The details are in the terms of service under “Processing on your behalf”.

On first contact the app registers your device with the server. For this it creates a random device identifier and a secret device key. Both values are completely random and allow no conclusions about the device you use. We store the device identifier, a digest of the key and the date of registration until we block the device or you ask us to delete them. For personal access, we create a user name and a password on your request. We store the user name and a digest of the password until you cancel the access.

The server logs sign-ins with the device identifier or the user name. On failed attempts it also logs the IP address. We use these logs to detect and investigate faults and unauthorized access. We delete them after 14 days. If we have to investigate a specific incident, we keep the entries it needs until the investigation is complete.

We are the controller for the device identifier, your credentials, the IP address and the sign-in logs. We process the device identifier or your credentials so that you can use the server. The legal basis is the performance of our contract with you (Art. 6(1)(b) GDPR). The sign-in logs help us run the server securely and reliably. For this we rely on our legitimate interest (Art. 6(1)(f) GDPR).

You can enter a different server. Its operator is then responsible for processing your data.

You are responsible for the audio and the text you transcribe through the server. For private conversations the GDPR usually does not apply (Art. 2(2)(c) GDPR). If you use the app for work, you need a legal basis of your own. This applies in particular to protected information under Art. 9 GDPR. Examples are information about health or religion. If you record other people, tell them beforehand. Also tell them when the audio is sent to a server.

Contact through the app

Under “Settings > Contact” you write us an email. The app only prepares the email and opens your email program. The email contains your subject and your text, your operating system, device model, app version and the version of the speech models. A diagnostic file is attached only if you switch that on.

We use your email only to answer your request. Afterwards we delete it. The legal basis is our legitimate interest in being able to help you and to improve the app (Art. 6(1)(f) GDPR).

Who else receives data

Both our servers are in Germany. The host is STRATO GmbH, Otto-Ostrowski-Straße 7, 10249 Berlin. There is a processing agreement with STRATO under Art. 28 GDPR. It is at www.strato.de/agb/avv (in German). For the audio and text we process on your behalf, STRATO is our sub-processor. How STRATO processes data and how long it keeps it is described in STRATO's privacy policy at www.strato.de/datenschutz (in German).

Beyond that we pass nothing on. Data does not leave the European Union. There is no automated decision-making and no profiling.

Your rights

Under the GDPR you have the following rights: access (Art. 15), rectification (Art. 16), erasure (Art. 17), restriction of processing (Art. 18), data portability (Art. 20) and objection (Art. 21).

We and our host store the credentials, server logs and contact emails described above. Our server processes audio and text only for transcription and translation and does not store them permanently. We cannot see or delete for you data that is only on your device. Write to us if you want to use one of these rights. We have to hand data over to third parties only when a law or an authority obliges us to (Art. 6(1)(c) GDPR).

You may complain to a data protection supervisory authority (Art. 77 GDPR). For example, the authority where you live. In Germany the federal commissioner lists all authorities. The list is at www.bfdi.bund.de.

Objection

If we process data on the basis of a legitimate interest, you can object on grounds relating to your particular situation. To do so, write to us at the address under “Who is responsible?”. We then stop processing the data, unless we can demonstrate compelling legitimate grounds or need the data to establish, exercise or defend legal claims.

Children

Children under 16 may use the app only with the consent of their parents or guardians. This also applies to our server. The app can also record children, for example in a conversation. Their data is then handled as described above.

Other privacy laws

If another data protection law applies to you, nothing changes for you. We collect only the data named above. We do not sell it or pass it on. Under those laws you have rights too, such as access and deletion. If you use these rights, we do not treat you any differently because of it.

Changes

This policy is part of the app. If it changes, you receive the new version with a new version of the app. The date above shows the current state.