GDPR and interview transcription

An interview recording is the most personal material in a study: a voice cannot be pseudonymised. Here is where Themera processes recordings, what you control and what to settle before uploading.

Try it free How it works

Transcribe 30 minutes free. No card needed.

Where your recording is processed

All details, including the other providers involved, are in the privacy policy.

Recordings are stored in a data centre in the EU.
They are transcribed by Mistral AI in Paris, also in the EU.
Mistral AI does not use the data to train models.
A data processing agreement (DPA) is available for organisations.

What you control

Delete the recording in one click once you have checked the transcript. The transcript stays.
Deleting a project deletes its recording as well.
Pseudonymisation replaces names, places and organisations with labels like [Person 1], and you review every match.
Real names are not stored. The mapping exists only in your browser, and you can download it.

Checklist before you upload

This is not legal advice. Settle these questions with your institution’s data protection officer or ethics committee.

Does participants’ consent cover recording, transcription by an external service and the retention period?
Is it clear who can access recordings and transcripts, and when they are deleted?
Are real names and the pseudonym mapping kept separately from the material?
Does your organisation need a DPA? Request it before the first upload.

Pseudonymised is not anonymous

A transcript without names can still point to one person: through their ward, their role, their years of service or one distinctive event. Under the GDPR such data is pseudonymised, not anonymous, and remains personal data. That is why the pseudonymisation suggests distinctive details too, not just names.