Trust · privacy
Privacy, in plain language
Designed for focused, responsible business use. Working principles while the full policy is finalized before broader invites:
01 · AudioWhere capture happens
- By default, live audio is transcribed using our cloud speech providers (Deepgram or OpenAI Whisper): audio is sent for transcription only and is not retained afterward.
- With Local mode, speech-to-text runs entirely on your device after the on-device model pack is downloaded, so raw audio never leaves it.
- Hybrid mode uses local speech-to-text when ready and falls back to cloud transcription automatically when it is not.
02 · AssistanceWhat help can use
- Assistance you request may send transcript text and selected workspace knowledge to the model provider you use. This is text, not audio, and applies regardless of STT mode.
- Cloud sync covers workspaces, session metadata, and notes you choose to save.
03 · ControlYour organization stays in charge
- The organization controls connected knowledge. Employees remain responsible for what they communicate.
- We do not use your conversation content to train Wisprline models.
- Contact hello@wisprline.com for deletion or export requests.
