Record yourself speaking English, see which sounds you got wrong, and get coached on how to fix them.
Pet project, for local use only.
make setup # first time only
make up # starts it at http://localhost:8501Needs Docker. Fill in .env (created by make setup) with:
AZURE_SPEECH_KEY+AZURE_SPEECH_REGION— required. Free F0 tier at Azure Speech (pick F0, not S0, which is not free).GEMINI_API_KEY— optional. Marks the passage up with how it should have been read; the coaching itself needs no key. Free at Google AI Studio.
make test # run tests, offline, no API calls
make check # lint + typecheck + test