Data Management

Last updated: May 15, 2026

LingoFlow documents implemented behavior separately from future improvements. Deletion is described only for data the product can currently delete.

Data type	Purpose	Retention approach	Deletion
Realtime audio stream	Transcription and translation processing	Not designed to be saved as history by default. Handled temporarily as needed for processing.	After the session ends, the product prioritizes not keeping the stream as saved history.
Recorded audio file	Recorded audio transcription processing	May be stored where needed for processing, retry, and result history.	User deletion or related job deletion is supported where implemented.
Transcript text	History, summarization, review	Saved when history/result storage is enabled.	The UI shows the scope that the user can delete.
Translation result	Translation history and review	Saved when history/result storage is enabled.	The UI shows the scope that the user can delete.
Summary	Reviewing conversations and recordings	Saved when history/result storage is enabled.	Deleted where related job deletion is implemented.
Usage logs	Billing, quota, abuse prevention	Stores seconds, status, billing state, and similar metadata. The design prioritizes not including text body or audio.	Retained for accounting, abuse prevention, and operations as needed.
Error logs	Incident investigation	The design prioritizes not including text body or audio.	Retained for operational needs.

AI training non-use

We do not use your audio, transcripts, translations, or summaries to train LingoFlow-owned AI models.

Free plan: server-side session and transcription result data is generally retained for 7 days where applicable.
Pro plan: server-side session and transcription result data is generally retained for 30 days where applicable.
Billing, quota, abuse-prevention, and operational logs may be retained separately.

TODO_SECURITY_VERIFY: confirm production log retention and Cloud Storage lifecycle settings before publishing stronger timing guarantees.