Handbook / Security and data protection
Data location
The cloud version of Universe Memory runs in Frankfurt am Main; speech recognition and the first language model process in the EU.
Purpose
Where data is stored determines law and trust. Universe Memory keeps knowledge, questions, processes, situation reports and drafts in an environment in Frankfurt am Main and uses services with processing in the EU first for speech and text.
How it works
- In the cloud version, the stored content of the organization is kept in Frankfurt am Main.
- Voice recordings are converted into text by a speech recognition service with a data center in the EU and are not stored by Universe Memory.
- Situation reports and drafts are created first at a language model with processing in the EU.
- Microphone only after your click
- Four-eyes principle
- Data in Frankfurt am Main
- Every approval in the log
Fig. 1 The three visibility scopes: the database does not deliver a restricted entry of another department; a department entry is seen by the department's members.
Language model as fallback
If the language model with processing in the EU cannot be reached, a second language model answers the request. It receives the same, narrowly selected evidence as the first model.
Tips
Before starting, clarify with your data protection officer which content belongs in “Restricted”.
Frequently asked questions
Where is the data of the cloud version stored?
In Frankfurt am Main.
Are voice recordings stored?
No. Universe Memory only stores the recognized text that you take over.
Related features
As of: · Concept & lead: Alexander Holtermann
