When we sit down with you – in a meeting room or on a video call – we take notes, and for some time now we have done so with technical support. This page explains what happens in the process and what explicitly does not, so that you can decide whether you agree to it.
In brief
- No audio is recorded. No audio file is created at any point.
- No voices are recognised or compared with one another.
- No contribution is attributed to any individual.
- Nothing leaves our devices. Speech recognition runs locally, not in a cloud.
- Without your consent, we switch it off. One person is enough.
- In video conferences, the conference audio is captured as well – with the same three constraints in place.
- The platform's recording indicator does not come on. That is why we tell you at the start.
Why we take notes
Meetings produce decisions, commitments and tasks. If nobody records them, they get renegotiated at the next meeting. If someone records them by hand, that person is busy with a notepad rather than with the conversation. Automated transcription is meant to resolve exactly this dilemma – nothing more.
It does not replace meeting notes. It provides the raw text from which we write the meeting notes that you then receive from us and can challenge. The record is the meeting notes, not the raw text.
Three built-in constraints that cannot be quietly switched off
No audio recording. The microphone signal is processed solely in working memory, where it is continuously overwritten. No audio file is created at any point – not even a temporary one. So there is nothing that could later be listened to, copied, passed on or requested by a third party.
No voice recognition. No voice profiles are created and no voices are matched, either within a meeting or across several meetings. The software cannot recognise anyone by their voice.
No attribution of contributions. The raw text carries no speaker labels – not even seemingly harmless ones such as “Speaker 1”. It is continuous text with no information about who said what.
These three points are not statements of intent but properties of the operating mode we use. We are not promising you that we will refrain from using a feature that exists – the feature does not exist. A promise can be withdrawn, reconfigured or simply forgotten; a missing feature cannot.
Nothing leaves our devices
Speech recognition runs locally on the consultant's computer. Neither audio nor text is transmitted to a cloud service or to a provider's language model. Our projects regularly deal with infrastructure, procurement, organisational questions and people. Material of that kind has no business passing through other people's data centres – whatever their terms of use may say. This applies to conference audio just as much as to the microphone.
In video conferences
With Teams, Webex or a browser-based conference, we also capture the conference audio, which includes the contributions of the other participants. The three constraints above apply unchanged: no audio file is created, no voices are recognised, and no contribution is attributed to any individual.
The platform's recording indicator does not come on. Teams, Zoom and Webex show all participants when a recording is being made within the platform. What we do is not a recording within the platform – so the indicator stays off. This is not a trick but a consequence of the technology; what matters to us is that you know about it. That is why we state explicitly at the start that notes are being taken, rather than relying on a signal that will not appear.
No screen content. macOS provides access to system audio through the same interface as screen recording – which is why the permission carries that name in the operating system. Only audio is captured. No screen content is stored, viewed or passed on.
Here, too, only with everyone's consent. Video conferences more often include people from other organisations who have never seen this page. That is why we explicitly ask everyone present at the start – a link alone is not enough for us.
What this explicitly does not mean
A promise that claims more than it delivers is worthless. Three limits are therefore part of it:
Content remains content. If a name is mentioned in the meeting, that name appears in the text. The transcript does not make the meeting anonymous; it refrains from attributing contributions to individuals.
Commitments keep their names. We continue to record tasks, commitments and dates together with the person responsible. That is the purpose of meeting notes and does not change – a task list without names would not be a task list.
In a small group, inferences remain possible. Someone who speaks in a meeting of five people may remain identifiable from the context, even without their name appearing in the text. The raw text is therefore free of attribution, but not anonymous in the legal sense.
What happens to the text
The raw text is used to produce the meeting notes. As soon as these have been agreed with you, we delete the raw text – no later than four weeks after the meeting. The meeting notes themselves are subject to the retention rules of the project in question.
If anything in the meeting notes seems to you to be wrong or abridged, tell us. We will correct it – as a matter of course for a machine transcript just as for handwritten notes, and, in our experience, just as often needed.
Your consent
One person who does not want it is enough – the software then stays off and we take notes by hand. No reasons required, no questions asked, no disadvantage to our working relationship.
You can withdraw your consent at any time, even in the middle of a meeting. It then ceases to apply from that moment; anything written up to that point remains lawfully collected and is handled according to the rules set out above. If you wish, we will instead delete the raw text immediately.
We record in the header of the meeting notes that consent was given. No further documentation is needed, and we collect nothing more for that purpose.
For agenda items concerning personnel matters, the assessment of individuals or anything expressly confidential, we switch off on our own initiative – without anyone having to ask.
Information pursuant to Art. 13 GDPR
Controller. WE SUCCESS Consulting GmbH, Wachaustr. 11, 70469 Stuttgart, represented by the Managing Director Andreas Pelzner. Email: info@we-success.de.
Purpose. Preparing the meeting notes and documenting our consulting services.
Legal basis. Art. 6(1)(a) GDPR – your consent. You may withdraw it at any time with effect for the future; this does not affect the lawfulness of processing carried out before withdrawal.
Data processed. The spoken word in text form, without attribution to individuals – in video conferences including the contributions of the other participants; in the meeting notes, additionally name and role in connection with commitments, tasks and dates.
Not processed. Audio recordings, voice characteristics, speaker profiles.
Recipients. None. The raw text does not leave our devices. The agreed meeting notes go to the distribution list agreed for the project.
Transfer to third countries. Does not take place.
Storage period. Raw text until the meeting notes have been agreed, and no longer than four weeks after the meeting. Meeting notes in accordance with the retention rules of the project in question.
Your rights. Access (Art. 15), rectification (Art. 16), erasure (Art. 17), restriction of processing (Art. 18), data portability (Art. 20) and objection to processing (Art. 21 GDPR).
Right to lodge a complaint. The State Commissioner for Data Protection and Freedom of Information of Baden-Württemberg (LfDI), Postfach 10 29 32, 70025 Stuttgart, poststelle@lfdi.bwl.de.
Automated decision-making. Does not take place. We read and edit the machine-generated text before any use.
Last updated
This page was last updated in October 2026. If anything changes in the operating mode described here, we will update this page – and tell you before we next transcribe one of your meetings.