Privacy Policy
How we collect, use, store, and protect your data when you use Mantrify.
Contact: constantine.ff@gmail.com
1. Introduction
Welcome to Mantrify. This Policy explains what data we collect, how we use it, and what rights you have. By using the Service, you agree to these terms.
2. Information We Collect
Account & identity
- Google Sign-in info: display name, avatar URL, email (for account creation and identification, not ads).
Content you provide
- Habit problem text (title, description, type).
- Mantra text (generated or edited).
- Feedback you submit.
Voice & audio
- Voice samples you upload/record for cloning.
- Generated audio (system TTS or cloned/self-recorded).
- Metadata (IDs, duration, language, related habit/mantra).
Safety & abuse signals
- High-risk content flags (e.g., self-harm/violence) detected via automated safety checks to block harmful requests and comply with legal duties.
- Minimal metadata for safety events (timestamp, request/result code, IP/UA/device fingerprint where applicable).
Logs & device info
- Access logs: timestamp, URL, browser/OS, coarse IP for city-level stats and security.
- App/LLM/voice logs: request IDs, latency, status codes, minimal de-identified context; not used for profiling or ads.
Cookies & similar tech
- Keep you signed in; remember locale/theme; measure traffic via GA/Clarity.
- No plain-text passwords or sensitive data stored in cookies.
3. How We Use Your Information
- Provide and maintain the service: accounts, mantra generation, voice generation/playback, delete/manage content.
- Improve product performance and UX through aggregated/anonymous analytics and error debugging.
- Security and abuse prevention: detect suspicious logins, abusive API use, and high-risk content; throttle or block when needed.
- Legal/compliance: respond to valid legal requests when required.
4. AI & Voice Notes
- Habit/mantra text is used only to generate guidance for your account; not to train public models.
- Voice samples/cloned voices are used only to generate/play audio for you; not reused for others or generic training.
- We send only minimum required data to Gemini 2.5 Pro, CosyVoice v3, and similar providers.
5. How We Share Information
Service providers (processors)
- Google OAuth (auth); Supabase/Postgres (data); Cloudflare R2 (audio storage); Gemini 2.5 Pro (mantra generation); CosyVoice v3 (voice); Resend (emails); Google Analytics, Microsoft Clarity (analytics).
Legal requirements & protection
- We may disclose data if required by law or to protect users/our rights/safety (including high-risk content cases), following proper legal process.
6. Storage & Retention
- Account/business data: kept while your account is active; deleting a habit/mantra hides it immediately and triggers DB + R2 audio cleanup within 24 hours (with retry/compensation on failures).
- Logs: access logs retained up to 30 days; LLM/voice logs up to 90 days.
- Backups: for disaster recovery only, deleted after retention.
7. Your Rights
Contact us to exercise your rights; we aim to respond within 24 hours on business days.
- Access and obtain a copy of your data (where applicable).
- Rectify inaccurate account info.
- Erase habits/mantras/audio in-product; request account/data deletion via email.
- Restrict processing or withdraw consent for non-essential features (e.g., analytics), where applicable.
- Data portability in jurisdictions that require it.
8. Children’s Privacy
Mantrify is for adults with full legal capacity. If under 18, use under guardian guidance. We do not knowingly market to children.
9. International Transfers
Some providers may process data outside your country. We work with vendors that follow strong security practices.
10. Changes
We may update this Policy; material changes will be communicated in-product or via email, with the updated date shown.
11. Contact
Questions or requests: constantine.ff@gmail.com. We respond within 24 hours on business days; your email is used only to handle your request.