Voice & Camera Data

This is the important part. Elovox does not create a voiceprint, a face template, or any other biometric identifier, and it does not try to recognize who is speaking or on camera. We turn on none of the speaker-identification features our transcription provider offers. Your recording is analyzed for how you came across — pace, clarity, expression — never to match you to an identity.

• Audio. Your browser records you and sends the audio to our server, which passes it straight to the transcription provider. It is not written to our storage, and we destroy our copy as soon as the transcript comes back — within the seconds it takes to process one recording.
• Camera frames (only if you turn the camera on). A small number of still frames — no more than twelve — are sent to the AI model to comment on your delivery, then discarded. They are never written to our storage.
• What we keep instead. The transcript, the metrics, and the coaching report are saved to your account, because that's what your progress is made of. The raw audio and the frames are not.

Recording only happens when you start a practice run, and the camera stays off unless you switch it on — video is never required to use Elovox. You can delete any session, or erase your whole account along with everything saved to it, from your account settings.

While your audio or frames are briefly with our transcription and AI providers to be processed, their own retention rules apply to the copy they hold. They are named, with links to their policies, in the Privacy Policy.

Email elovox.app@gmail.com with anything about your voice or camera data and we'll respond within 30 days.