ChinSpeak Privacy Policy
A plain-language explanation of what ChinSpeak processes, why it is needed, where it is kept, and the choices available to you.
Effective
1 September 2026
Account
Not required
Advertising
No ads or tracking
Normal speech content is transient: recordings stay in memory for seconds, transcripts and translations are not stored, synthesis text is cleared after the render, and generated audio is not retained by the public API. A feedback report you choose to send is the deliberate exception explained below.
Who operates ChinSpeak
ChinSpeak is a mobile application developed as a joint collaboration between Chinland InfoTech and Chin Community, USA. Chinland InfoTech operates the app and its online services and is the contact responsible for this policy.
This policy applies to the ChinSpeak mobile app and the services used to recognise, translate, and synthesise speech. It does not apply to unrelated third-party websites you may visit.
Data the app processes
- Voice recordings. When you tap the microphone, the app records the short clip you choose to provide and sends it to the Chin Voice Lab service for speech recognition.
- Recognised and translated text. The speech result, any correction you make, and the requested language direction are processed to produce a translation. Text may be sent to our Chin Translator service.
- Listen requests. When you tap Listen, the translated text, language, and selected voice are processed to generate playable audio.
- App-install information. The app creates a random identifier for that installation and sends the device platform so the service can issue an access token, apply limits, and protect the API. It is not an advertising identifier.
- Feedback you submit. An issue report may include its category, description, language direction, source text, corrected text, translation, request identifier, app version, and platform. Your recording is attached only if you separately turn on the audio option before sending.
- Security information. The service processes IP addresses for short-lived rate limiting and may create normal infrastructure access or error logs.
- Data stored on your device. Preferences and translation history are stored locally. Recent unsaved history is bounded; phrases you explicitly save remain until you remove them, clear history, or uninstall the app.
Why we use this data
We process this data to:
- recognise speech, translate it, and generate audio playback;
- maintain language, history, and playback preferences on your device;
- authenticate the app installation, prevent misuse, and keep the service reliable;
- investigate translation problems and respond to reports you choose to send; and
- improve Chin-language quality when you explicitly provide corrections or consented feedback audio.
Where local law requires a legal basis, these activities are performed to provide the service you request, for our legitimate interests in security and product quality, or with your consent for optional feedback audio.
Where data is processed
ChinSpeak sends data over encrypted HTTPS connections to Chinland InfoTech's Chin Voice Lab and Chin Translator services. We also use cloud hosting, database, queue, and object-storage providers acting on our behalf to run those services. Ordinary speech recordings and public synthesis output are not written to object storage; object storage may be used only when you explicitly attach a recording to a feedback report.
We may disclose information if required by law, to protect users or the service, or as part of a business reorganisation with appropriate safeguards. We do not sell or rent voice recordings, translations, or device identifiers.
Depending on where you live and where our service providers operate, processing may occur outside your country.
Retention and caching
- Speech recordings are transient. Audio sent to the normal recognition and translation endpoints is held in memory only, never written to disk or object storage, and discarded when recognition finishes. It normally exists for seconds and expires within five minutes even if processing fails.
- Transcripts and translations are not stored by the public speech API. They are computed, returned to the app, and forgotten. No database row contains what you said or its translation.
- Synthesis content is request-only. Text sent when you tap Listen exists on the job only while the audio is rendered and is cleared whether rendering succeeds or fails. Audio returned by the public API is not retained. The same rule applies to Falam and English.
- ChinSpeak may keep the current generated clip temporarily in app memory so you can tap Listen again without another request. That in-app copy is cleared when the translation, language, or active result changes, and it does not become History.
- Local history and preferences remain on your device until you clear them, disable new history, remove individual items, or uninstall the app.
- The server may keep the random app-install identifier and content-free operational job metadata as needed to authenticate, secure, diagnose, and operate the service. This metadata does not include the transcript or translation.
- Feedback is the exception. An issue report deliberately retains the report text and, only if you separately opt in, the attached recording so a reviewer can rule on it. Attached report audio is deleted automatically 90 days after it is uploaded, by a retention rule on the storage bucket that holds it — you do not have to ask. You may still request earlier deletion using the contact below.
- Short-lived IP rate-limit records expire with the protection window. Infrastructure providers may retain access or security logs under their own operational schedules.
Your choices and deletion requests
You can decline microphone access, stop a recording, edit recognised text before translating, turn local history off, clear local history, and keep feedback audio switched off. Declining microphone access prevents voice capture but does not affect access to this policy.
Normal speech audio, transcripts, translations, synthesis text, and generated audio are not available for later deletion because the public API does not retain them. To ask about a feedback report or other retained operational records, request deletion, withdraw optional feedback consent, or raise a privacy concern, email admin@chinlandinfotech.com. Include “ChinSpeak privacy request” and enough information to help us locate the relevant request or report, but do not send additional sensitive content.
Your rights may vary by location. We may need to verify a request before acting on it, and we may retain limited records where required for security, fraud prevention, or legal obligations.
Security
We use encrypted transport, scoped app credentials, access controls, and operational safeguards intended to protect data. No system can guarantee absolute security, so we also avoid persisting ordinary speech content and do not require a user account.
Children
ChinSpeak is not directed to children under 13. If you believe a child has submitted retained personal data without appropriate permission, contact us so we can review and delete it where required.
Policy changes
We may update this policy when the app, service providers, or retention practices change. The effective date at the top will be revised, and material changes may also be explained in the app or release notes.