Step-by-step instructions
Navigate to Anonymisation from main navigation
Click "Upload Document" or drag-and-drop file (PDF, DOCX, TXT, images)
Upload modal appears with file selected
Expand "Anonymization Settings" collapsible section
Select categories to anonymize (default: workspace defaults):
All Variables (overrides all others - anonymizes everything)
Names (people, signatures)
Addresses (street addresses, postal codes)
Emails (email addresses)
Phone Numbers (mobile, landline)
Bank Details (IBANs, credit card numbers)
ID Numbers (national ID, passport numbers)
Dates (birthdays, deed dates, timestamps)
Prices & Amounts (currency values)
Company Names (business entities)
URLs & Domains (websites, domains)
Custom Terms (user-defined terms to anonymize)
Click "Anonymize" button
Status shows "Processing" (typically 30–60 seconds)
Status changes to "Completed"
Click document to review side-by-side comparison:
Left pane: Original document
Right pane: Anonymized version with highlighted replacements
Verify accuracy of anonymization
Click "Download Anonymized" to save redacted version (PDF or DOCX)
Practical example
Notary Sophie Mercier needs to share a 45-page liquidation deed of "SA Dupont Industries" with Brussels law school for teaching case studies. She uploads the deed to Anonymisation, selects categories: Names, Addresses, IBANs, ID Numbers, Phone, and Email (unchecks Dates and Company Names because students need to understand timeline and company structure). Processing completes in 45 seconds. She reviews the result: all client names replaced with [PERSON_1], [PERSON_2], all addresses with [ADDRESS_1], all IBANs with [IBAN_1], but dates (deed signing date, liquidation timeline) and company name "SA Dupont Industries" preserved. The anonymized deed is now GDPR-compliant for external academic sharing while maintaining legal structure for educational value. She downloads the PDF and emails it to the law professor.
Notes and limitations
AI accuracy: Approximately 95% detection accuracy – always review results for critical documents
Supported formats: PDF, DOC, DOCX, TXT, images (JPG, PNG, GIF)
Processing time: Small documents (<10 pages): 30–45 seconds; large documents (100+ pages): 5–15 minutes
Category configuration: Select categories before processing – cannot undo specific categories after anonymization (must re-process)
OCR for scanned PDFs: Images and scanned PDFs automatically OCR'd before anonymization
Language support: Works best with French, Dutch, English documents
Multi-format export: Download as PDF (preserves formatting) or DOCX (editable)
FAQ
Q: Is anonymisation GDPR compliant for sharing documents externally?
A: Yes, the Anonymisation module is specifically designed for GDPR compliance with Belgian data protection requirements. It detects and redacts personal data across 12 categories including names, addresses, IBANs, ID numbers, and more. However, always review the anonymized document before sharing externally to ensure all sensitive information is properly redacted, as AI detection is approximately 95% accurate (not 100%).
Q: Can I undo anonymisation and restore the original document?
A: The original document is always preserved in a separate storage bucket and never deleted. You cannot "undo" anonymisation on the anonymized version, but you can always access the original. If you need to re-anonymize with different settings (e.g., exclude Dates category), click the configuration icon, update settings, and click "Re-anonymize" to generate a new version with different categories applied.
Q: What file formats are supported for anonymisation?
A: Supported formats include PDF, DOCX, TXT, and images (JPG, PNG, GIF). Legacy .doc files are not supported. For scanned PDFs or images, the system automatically performs OCR (Optical Character Recognition) to extract text before anonymization, though OCR quality depends on scan resolution and clarity. Maximum file size is 50MB.