On-device first
The app supports English speech only. Recognition runs on your Mac when the system supports on-device recognition for English. Otherwise Apple's speech service may process your audio.
Security
Clear about speech and data.
SAYA processes speech to follow a script and provide delivery feedback. These are the current processing options and recording defaults in the app.
Audio recording off by default
SAYA does not record raw audio by default. It stores recognized text and session events locally for session review and diagnostics.
No biometric profiling
SAYA does not infer gaze, emotion, stress or identity. Pro and Team feedback includes estimated speaking pace and long pauses. Recognition errors can affect these figures.