Last updated: July 28, 2026
What FluentMeet processes
FluentMeet processes audio in real time to provide speech recognition, translation, and translated voice. The commercial service also processes minimal device authorization, session, usage, performance, and security metadata.
What we do not retain by default
FluentMeet does not retain raw meeting audio, translated audio, full source transcripts, full translated transcripts, meeting URLs, participant lists, or meeting titles by default.
Account and device data
The Closed Beta uses invitation and device-bound authorization. The macOS app stores its device credentials in secure operating-system storage. The Cloud Portal does not collect or display Device Tokens, Session Grants, invite codes, or provider credentials.
Usage and reliability data
We may retain session duration, language direction, provider-neutral usage totals, latency, safe error codes, application version, device architecture, and security events. These records help enforce beta limits, diagnose reliability, and control cost.
AI providers
Realtime media may be processed by an approved AI provider selected by FluentMeet infrastructure. Provider credentials and provider-specific details are not exposed to the client. Provider use is governed by regional, security, and retention controls.
Your choices
You can stop a session at any time, deactivate FluentMeet, and restore your original audio devices. Future account deletion and export controls will be introduced before public account launch.
Security
FluentMeet uses short-lived authorization, one-time realtime grants, server-side provider credentials, structured safe logging, and access controls. No online system is risk-free; security concerns should be reported through the Closed Beta support channel.
Contact
A public privacy contact will be published before external distribution. During the invite beta, use the support contact provided with your invitation.