Pseudonymization
Replace data with pseudonyms that are consistent and reversible within the session.
Pseudonymization replaces each detected data point with a consistent pseudonym that remains consistent throughout a document or corpus of documents: all syntactic versions of the same entity are assigned the same pseudonym.
Unlike anonymization, it is reversible thanks to restoration: the data is transformed but remains re-identifiable via a matching key.
Pseudonymization is one of the measures recommended by the GDPR to limit the risks associated with the processing of personal data.
Usage
- Open the text editor from the Plain Text page or the file processing tool from the Sessions page
- Select “pseudonymization” as the processing type from the session’s Overview tab
- Paste your content or import your files
How It Works
When you start a processing session in pseudonymization mode, the technology generates a unique pseudonym for each detected entity and saves it to the pseudonymization table for the current session, accessible from the Pseudonyms tab:
| Data | Associated Pseudonyms |
|---|---|
| Marie Dupont, Marie, Mrs. D. | PR1a, PR1b, PR1c |
| Thomas M., Thomas Martin, T. Martin | PR2a, PR2b, PR2c |
| Technoplus SAS, TECHNOPLUS | CIE1a, CIE1b |
| Vival Group, VIVAL GROUP, VIVAL | CIE2a, CIE2b, CIE2c |
The pseudonym (PR1) identifies the entity, and its closing letter the exact spelling written at that point of the document: PR1a is "Marie Dupont", PR1b is "Marie". A value shared by two entities, such as a family name, carries both numbers and the letter of the last one (PR1/2a).
This table is retained throughout the session, which allows for:
- Consistency of pseudonyms within a single document
- Sharing pseudonyms across multiple documents within the same session
- Subsequent restoration of the original values in place of the pseudonyms
When to Use Pseudonymization?
- Working with an AI assistant to draft, summarize, or analyze
- Leveraging your internal document assets (knowledge base, RAG)
- Collaborating with a service provider who needs to understand the document’s logic
Restoration
The pseudonymization mode is always integrated with restoration. As long as the session is not deleted (automatically or manually), restoration allows you to retrieve the original values from pseudonymized content.
In practice: Thanks to restoration, a pseudonymized document instantly becomes usable. Whether it has been edited by a colleague, supplemented by a third party, or enriched by AI, you retrieve your final document with the original data, ready for use.
Retention of Pseudonyms
Did you know that you can configure how long pseudonyms are retained? Go directly to the Privacy page to set your preferences.
By default, documents within a session are retained only for the duration of processing (strict non-retention policy).
Sessions and their associated pseudonyms are automatically deleted based on your selected settings. Of course, you can also manually delete sessions directly from the Sessions page. This page also allows you to view all your active sessions.