How nthoughts handles your voice and your data

Last updated: August 10, 2026. Private, cloud-secured, and local-first by design.

1. Information We Collect

We collect minimal information to deliver personalized coaching: your Google profile data (name, email, and avatar) for authentication, session speech-to-text transcriptions, and local configuration preferences.

2. How We Process Speaking Data

Your verbal transcripts are analyzed securely by LLM inference engines to generate grammar corrections. Audio output is synthesized using high-fidelity ElevenLabs voices. We do not store raw voice recordings indefinitely on our servers.

3. Local Cache & Spaced Repetition

We believe in a local-first architecture. Any vocabulary mistakes marked for spacing and recall review are cached securely in your browser's local storage (localStorage) under 'nthoughts_mastered_mistakes'.

4. Token & Security Protection

Your authentication tokens are stored securely in browser cache and transmitted over encrypted SSL channels. We do not sell, rent, or lease your profile metadata or learning transcripts to third-party ad networks.

5. Data Export & Deletion Rights

You retain complete ownership over your learning history. At any time, you can visit the Settings panel to export your compiled curriculum history or permanently request the purging of your practice sessions.

6. Third-Party Advertising & Google AdSense

We use third-party advertising vendors, including Google AdSense, to serve ads when you visit our website. Google uses cookies (such as DART cookies) to serve ads based on your prior visits to our site and other websites on the Internet. You may opt out of personalized advertising by visiting Google Ads Settings (https://adssettings.google.com) or AboutAds (http://www.aboutads.info).

Your security and calm learning process are our primary design parameters.

Ready to speak with confidence?

Join the beta and start practising with Maya today. No card required.

Start speaking free

No card required · Free during launch