Step-by-step instructions
Upload and process document through Anonymisation (see "How to Use Anonymisation")
Review anonymized result and identify terms the AI missed
You can then proceed in two ways:
Method 1: Settings Method: In document details view, click "Sliders" icon or "Configuration" button
Category configuration dialog opens
Scroll to "Custom Terms" section
Click "Add Custom Term"
Enter exact term to anonymize:
Company-specific: "Dupont Industries", "Project Phoenix"
Location-specific: "Brussels office", "Wallonia branch"
Sensitive codenames: "Operation Blue Sky", "Deal Tango"
Repeat for all additional terms
Text Highlighting Method: within the left pane, highlight any term which should also be anonymised, and click on "add sensitive term". This will then apply to all the occasions that this term appears in the text.
Once you have made your modifications:
Click "Re-anonymize"
System regenerates document with custom terms replaced
New version created preserving original and previous anonymized versions
Review updated anonymization
Download if satisfactory by clicking on "Export".
Practical example
Notary Vincent Laurent anonymizes a merger agreement mentioning "Project Phoenix" (internal codename for the merger) and "Dupont Industries Brussels headquarters" (specific office location). The AI successfully anonymizes personal names and IBANs but misses "Project Phoenix" and partially anonymizes the company office location. He clicks the configuration icon, adds custom terms: "Project Phoenix" and "Brussels headquarters" (keeps "Dupont Industries" since company name should remain), selects "Custom Terms" category, and clicks "Re-anonymize". The system regenerates the document in 40 seconds with "Project Phoenix" and " Brussels headquarters" replaced by "XXX" while preserving the company name as requested. The document is now fully anonymized for sharing with external legal counsel.
Notes and limitations
Exact term matching: Custom terms must match exactly (case-sensitive by default)
Re-processing required: Adding custom terms requires re-anonymization (generates new version)
Per-document configuration: Custom terms apply only to current document, not saved globally
Workspace defaults don't include custom: Workspace default categories don't include custom terms (must add per document)
No term library: Cannot save frequently-used custom terms for reuse across documents
Version history: Each re-anonymization creates new version (original always preserved)
No spell-check: Typos in custom terms won't match document text (ensure spelling exact)
FAQ
Q: Can I add custom terms before processing or must I re-process the document?
A: You can add custom terms before initial processing by expanding "Anonymization Settings" during upload and adding them in the "Custom Terms" field. However, if you discover missed terms after processing, you must configure them and re-anonymize. Re-anonymization is fast (30–60 seconds typically) and preserves all previous versions.
Q: Are custom terms saved for future use on other documents?
A: No, custom terms are per-document configuration only and are not saved globally or reused automatically on other documents. Each document requires manual addition of custom terms. This limitation may be addressed in future updates with a workspace-level custom terms library. For now, keep a text file of common custom terms to copy-paste when needed.
Q: Can I see what will be anonymised before processing (preview mode)?
A: No preview mode is available currently. The workflow is: (1) process document, (2) review results, (3) add custom terms if needed, (4) re-process. To minimize re-processing, consider creating a checklist of your office's common custom terms (project names, office locations, sensitive terminology) and add them during initial upload if known.