Skip to main content

How your data is used to improve model performance

Learn how your data helps us improve accuracy, fully optional, and under your control

Simon avatar
Written by Simon
Updated this week

Our transcription and subtitling models get better with real-world examples. When you choose to share your files, you’re helping them learn to handle new accents, languages, and recording conditions, making results more accurate for everyone.

What we use your data for

If you enable data sharing, we, and our trusted partners, may use your uploaded files to:

  • Train and improve our transcription models

  • Test accuracy across different accents and languages

  • Help the models adapt to various recording environments and audio setups

Your data is never used for anything else.


You're in control of your data

Participation is optional. You can disable data usage for model training at any time by:

  • Click Settings and members

  • Click Workspace

  • Scroll down to Data Usage

  • Select Don't use my data

Opting out will not affect the availability or accuracy of your transcriptions.


Enterprise Privacy

As an Enterprise organization, you are always opted out by default. Just like any other user, you can choose to help improve AI models through the same workspace settings if you wish to contribute to better transcription technology.

If you wish to opt in:

  • Click Settings and members

  • Click Workspace

  • Scroll down to Data Usage

  • Select Help improve AI

Did this answer your question?