Think'It Gallery
Privacy Policy — Think'It Gallery
DRAFT — for review by a qualified lawyer before publication. Prepared 11 October 2026. Not legal advice.
Online: https://gallery.shaneteks.dev/privacy.html
Developer: Richard van Wyk, an individual developer in Namibia ("we", "us").
Contact: richardshanemccaw@gmail.com
Applies to: the Think'It Gallery app for Android (Google Play and direct-download versions).
The short version
- Think'It Gallery has no ads and works offline. It organises, searches and edits the photos, videos, audio and documents on your phone, and all processing happens on your phone.
- We do not collect, store, sell or share your personal data. We have no servers that receive your files, searches, or usage.
- No cloud AI. Every AI feature (search, voice search, transcripts, text in photos, cut-outs, People) runs on your phone. The AI models are optional downloads you choose.
- No account, no advertising, no analytics, no tracking SDKs, no in-app purchases. The app never syncs your files with any server.
- The app connects to the internet only to download the AI models you choose (search, transcripts, text in photos, cut-outs, people; each once) and, in the direct-download version, to check for and download app updates.
1. What the app accesses on your phone
| What | Why | Where it goes |
|---|---|---|
| Photos, videos, audio, documents you allow it to see | Show, search, edit, share, rename, delete them at your request | Stays on your phone |
| AI "embeddings" (number lists describing each file) and a search index | Make search by description work | Stored privately inside the app on your phone, never backed up or copied to a new phone (rebuilt there); deleted when you uninstall |
| Transcripts of audio and video, and text read from photos (only when you ask for them or switch them on) | Let you read, copy and search what was said or written | Stored privately inside the app on your phone, never backed up or copied to a new phone (rebuilt there); deleted when you remove them or uninstall |
| Settings (theme, grid size, which tips you've seen, chosen AI model) | Remember your preferences | Stored privately on your phone |
| Faces in your photos and the names you give them (only if you use People) | Find photos of someone by name | Stored privately inside the app on your phone, never backed up; deleted when you delete it in Settings › People or uninstall |
| "Recently replaced" originals | Let you undo an edit that replaced a file, for 24 hours | Private app storage on your phone; deleted automatically after 24 hours |
| Basic device facts (RAM, free storage, chip, Android version) | Recommend an AI model your phone runs well | Read on the phone; never sent anywhere |
We process this data only on your device. It is never transmitted to us or to anyone else by the app.
Face data (People)
People is optional and off until you download its small face model and name someone.
- On this phone only. Faces are found and matched on your phone. Think'It never sends your photos or face data anywhere; we never receive it.
- Names are yours. A name exists only because you typed it. The app never suggests or looks up who someone is.
- What is stored: for each photo looked at, the position of each face and a list of numbers describing it (a "face embedding"), plus the names you created. It is kept in the app's private storage and excluded from Android backups and device transfers.
- Delete anytime: Settings › AI features › People lets you rename or delete a person (their face data goes too) or delete all face data at once. Uninstalling the app deletes it as well. Your photos are never changed.
- No sharing: face data is never shared, sold or used for anything else. The app only connects to the internet to check for updates and to download the models you choose.
2. Network connections
1. AI model downloads — when you choose a model, the app downloads its files (about 5–500 MB) from our public GitHub Releases page (github.com, operated by GitHub, Inc.). If a file isn't available there, the app downloads the identical file from the original publisher's page instead: transcript models from the whisper.cpp project on Hugging Face (huggingface.co, Hugging Face, Inc.), text-reading models from the RapidOCR project on ModelScope (modelscope.cn, Alibaba Group), cut-out models from the rembg project on GitHub and the MobileSAM files on Hugging Face, face models from the OpenCV Zoo project on Hugging Face. The app checks every file's SHA-256 fingerprint before use. As with any download, the host receives your IP address and standard request headers; see its privacy statement. No information about your files, searches, transcripts or photos is sent. Transcribing, reading text and cut-outs run entirely on your phone.
2. App updates (direct-download version only) — the app fetches a small update file from GitHub (when you open it, at most twice a day; once a day in the background if "Notify me of updates" is on; and when you tap Check for updates) and, if you agree, downloads the new version. The Google Play version never downloads or installs app files itself: it is updated only through Google Play.
3. Sharing — when you share a file, Android sends it to the app you pick. That app's own privacy policy applies.
3. Permissions
Photos and videos / Music and audio / (optionally) chosen document folders, Notifications (progress of indexing and new-media syncing), Set wallpaper, Modify system settings (only when you choose "Set as ringtone"; you grant it in Android's settings), Foreground service (see below), Run at startup / background work (to index new media while the app is closed, if you enable auto-sync), Internet (the downloads above only), Microphone (only while you use voice search after tapping the mic: on Android 12+ speech is recognised by your phone's on-device recogniser; on older phones the app asks Android's recogniser for offline mode. The app records nothing, stores no audio and sends nothing).
Foreground services run only with a visible notification: data sync while an AI model download or library indexing that you started is in progress, and media playback while music keeps playing after you leave the player. The direct-download version may also request "Install unknown apps" (updates) and "All files access" (documents). You can revoke any permission in Android settings; the related feature then stops working, the rest of the app keeps working.
4. Children
Think'It Gallery is a general-audience utility and is not directed at children under 13. It collects no data from anyone.
5. Your rights (worldwide)
Because we hold no personal data about you, there is nothing for us to access, correct, export or erase on our side. You control everything on your device: delete files in the app, clear app data in Android settings, or uninstall the app to remove the index, models and settings. If you contact us by email, we use your message only to reply and delete it when the matter is closed.
This approach is designed to be consistent with privacy principles under:
- Namibia — Article 13 of the Namibian Constitution (right to privacy). As of October 2026 Namibia's Data Protection Bill has not been enacted and there is no data-protection authority; we will update this policy if it becomes law.
- EU/UK GDPR — no personal data is processed by us as controller; on-device processing happens under your control.
- South Africa POPIA, California CCPA/CPRA and similar laws — we do not collect, sell or share personal information.
6. Security
Your data stays on your device and is protected by Android's app sandbox and your screen lock. Downloads are fetched over HTTPS and integrity-checked (SHA-256).
7. Changes
We will post changes on this page and update the date above. Material changes will also be noted in the app's release notes.
8. Contact
Richard van Wyk · richardshanemccaw@gmail.com · Namibia