Ad-free Android · AI · com.michael.privai
PrivAI
Ad-free privacy AI workspace for Android. Secure notes, on-device voice transcription, document OCR (ML Kit/Tesseract), and local summaries — no ads, no content cloud uploads.
Completely ad-free — no ads, no trackers, no sponsored clutter.
Ad-free Android app
What is PrivAI?
PrivAI is an ad-free Android workspace for secure notes, on-device voice transcription, and offline document OCR — core processing stays local with no content cloud uploads and no ads.
Ad-free advantage: Completely ad-free — no ads, no trackers, no sponsored clutter. MichaelSam94 apps are built without banner ads, interstitial ads, or video ad interruptions.
Key capabilities
- Local workspace notes with tags and search
- On-device speech transcription
- Photo OCR with ML Kit or Tesseract
- Local bullet summaries and keyword highlights
- Completely ad-free privacy workflow
Frequently asked questions
Is PrivAI ad-free?
Yes. PrivAI is completely ad-free. Your notes and transcripts are not monetized with advertising.
Does PrivAI upload my notes or photos?
Core notes, transcripts, and OCR extracts are stored locally. OCR images are read when you pick them and are not uploaded as a cloud workflow.
Can PrivAI do OCR in Arabic?
Yes. English and Arabic tessdata are bundled for high-accuracy Tesseract OCR alongside ML Kit fast mode.
PrivAI is an ad-free Android workspace for secure notes, on-device voice transcription, and offline document OCR — core processing stays local with no content cloud uploads and no ads.
Ad-free: Completely ad-free — no ads, no trackers, no sponsored clutter.
Secure notes, voice transcription, and document OCR — 100% on-device.
PrivAI is an Android workspace for capturing ideas, transcribing speech, and extracting text from photos. Core processing runs locally on your phone. No account required. No cloud uploads for your content.
Features
- Workspace notes — Create, tag, search, and open detailed note views. Everything stays in a local Room database.
- Voice transcription — Record and transcribe speech using the device speech recognizer, processed on-device.
- Document OCR — Pick a photo from the system image picker and extract text offline with ML Kit (fast mode) or Tesseract (high-accuracy mode). English and Arabic tessdata are bundled.
- On-device summaries — Bullet points, keyword highlights, and simple sentiment hints from local NLP — no API keys needed for basic intelligence.
- Unified dashboard — Switch between Notes, Audio Transcripts, and OCR Extracts from one home screen.
Privacy
| Data | Handling |
|---|---|
| Notes & transcripts | Stored locally in Room |
| OCR input images | Read only when you pick one; briefly cached in app storage for processing |
| OCR output | Extracted text saved locally; images are not uploaded |
| Summaries | Computed on-device via LocalTextIntelligence |
| Accounts / cloud sync | Not used |
Tech stack
- UI — Jetpack Compose, Material 3, Navigation Compose
- Architecture — ViewModel, Kotlin Coroutines, StateFlow
- Storage — Room, DataStore (OCR preferences)
- Speech — Android
SpeechRecognizerviaAndroidSpeechTranscriber - OCR — ML Kit Text Recognition, Tesseract4Android
- Testing — JUnit, Robolectric, Roborazzi (Play Store screenshots)
Permissions
| Permission | Purpose |
|---|---|
RECORD_AUDIO | Live voice transcription |
INTERNET | Required by some on-device ML libraries; user content is not uploaded |
OCR uses the system photo picker (GetContent); no storage or media-library permissions are declared.
Play Store assets
./gradlew generatePlayStoreAssets
Tests
# Unit tests (excludes Play Store screenshot tests)
./gradlew :app:testDebugUnitTest
# Play Store screenshot generation (Roborazzi)
./gradlew generatePlayStoreAssets
