Cantonese Learner has no accounts, no sign-up, no advertising and no tracking cookies. This page says what actually happens to data rather than what would be legally permissible.
Cantonese Learner is a browser app served by GitHub Pages. Lessons are built ahead of time by a scheduled job, not while you use it.
No account, no sign-up and no tracking cookies. Your settings and progress live in your browser's local storage and are never uploaded.
Speech grading uses your browser's built-in speech recognition (the Web Speech API, available in Chrome). In Chrome, that recognition runs on Google's servers: while you hold the record button, your audio is sent to Google, which returns the recognized text. This site never receives, stores or sees the audio or the text; the comparison against the lesson line happens in your browser. Google's handling of that audio is covered by Google's own privacy policy. The browser asks your permission the first time, and listening and reading keep working if you decline.
Fonts are served from this site, so no font provider is contacted and no content delivery network is involved. There is no analytics provider of any kind.
This project is aimed at adults and students in higher education. It is not directed at children under 13 and collects nothing that would identify one.
Material changes appear here with a new date at the top. Questions go to lyhjeremy@gmail.com.