Privacy Policy
At Reader AI (“we”, “us”, or “our”), we are committed to safeguarding the privacy and personal data of students, researchers, professionals, and avid readers. This Privacy Policy outlines how Reader AI (the AI-powered audio reading application accessible at https://readerai.edgarodey.com) handles your data and neural speech synthesis.
Core Privacy Commitment: Account-Based AI Narration with Local File Storage
Reader AI is an intelligent speech application designed to convert study materials and documents into natural spoken audio. Full original files (PDFs, EPUBs, Word documents, PowerPoint slides) are parsed and stored locally in your browser's IndexedDB on your device. Only small reading text segments are transmitted across secured TLS connections to our AI speech microservices for voice synthesis. We never sell your personal data or train public AI models on your private documents.
1. Identification of Application and Creator
Application: Reader AI — Intelligent Neural Audio Reading Platform.
Developer: Edgar Odey, software engineer and creator of AI academic technologies.
Developer Website: https://edgarodey.com
Contact Email: eo@edgarodey.com
2. User Accounts & Mandatory Authentication
Reader AI requires all users to maintain a registered account to access the platform. Authentication is managed securely through Google OAuth 2.0 powered by Supabase:
- Information Collected upon Sign-In: Your name, email address, profile picture URL, and unique user identifier provided by Google.
- Purpose of Account: To authenticate access to AI voice generation services, prevent unauthorized automated abuse of GPU inference servers, and synchronize your document library metadata and reading progress across your desktop, phone, and tablet.
3. Document Ingestion & AI Speech Synthesis
A. Document Parsing & Local Storage
When you upload a document (PDF, EPUB, Word .docx, PowerPoint .pptx, image notes, or text files), parsing occurs locally using client-side libraries (PDF.js, Mammoth.js, JSZip, Tesseract.js OCR). The full source file is stored exclusively in your browser's IndexedDB database (ReaderStorageDB). Your original full-length files are not permanently stored on remote database servers.
B. AI Voice Inference (Kokoro-82M On-Device Neural TTS)
When you listen to a document using Kokoro-82M TTS, audio synthesis runs 100% locally on your device inside your browser (via WebGPU or WASM). Model weights are downloaded once and cached permanently in your browser's offline cache. Your document contents are never transmitted to external servers for speech synthesis.
C. On-Device Fallback (Browser Web Speech API)
If you select the Browser Web Speech API in Settings, audio synthesis executes 100% on your local device hardware using your operating system's native speech engine without network data transmission.
4. Google API Services User Data Policy Compliance
Reader AI strictly complies with the Google API Services User Data Policy, including the Limited Use requirements:
“Reader AI's use and transfer of information received from Google APIs to any other app will adhere to the Google API Services User Data Policy, including the Limited Use requirements.”
- No Transfer to Third Parties: We never sell, transfer, or license user data received through Google OAuth to external data brokers or ad networks.
- No Advertising Use: We never use user information for commercial advertising, retargeting, or profiling.
- No AI/ML Model Training: We explicitly do NOT use Google user data or uploaded user documents to train, retrain, or improve generalized foundation AI or machine learning models.
- Human Inspection Prohibited: No human personnel read or review user files, documents, or personal credentials unless explicitly consented to for support or required by lawful subpoena.
5. Data Storage, Cloud Synchronization, and Deletion
Reader AI employs a hybrid architecture designed for performance and speed:
- Local IndexedDB (
ReaderStorageDB): Caches full document files, extracted text blocks, and OCR caches on your local machine for rapid offline loading. - Supabase Cloud Database: Synchronizes document metadata (title, page count, progress percentage, last opened timestamp) and bookmarks under your authenticated user ID so you can transition between desktop and mobile devices.
Delete Individual Documents
In your Library, click the delete icon on any document card to immediately remove the file and extracted blocks from local IndexedDB and cloud sync records.
Wipe All Local Cache
Navigate to Settings and select “Clear Local Data” to permanently remove all stored PDFs, book records, and audio progress caches from your current browser.
6. Security Standards
We implement comprehensive industry-standard security safeguards:
- TLS/SSL Encryption: All communications between your browser, our Next.js application servers, and the AI speech inference cluster are strictly encrypted in transit.
- Row-Level Security (RLS): Supabase cloud databases utilize database-enforced Row Level Security policies, ensuring users can only read and write their own documents and sync records.
- Browser Sandboxing: Reader AI operates within standard modern web browser sandboxes to prevent cross-origin data exposure.
7. Updates to This Privacy Policy
We may update this Privacy Policy from time to time to reflect modifications in our AI speech capabilities, infrastructure, or regulatory compliance. Any updates will be posted on this page with an updated “Last Updated” timestamp.
8. Contact Information & Privacy Inquiries
For any inquiries regarding this Privacy Policy, your account data, or our AI speech architecture, please contact the developer:
Edgar Odey (Lead Developer & Creator of Reader AI)
Email: eo@edgarodey.com