Automation
Presidio is an open-source toolkit maintained by the Data Privacy Stack community. It combines named entity recognition with rules to flag potentially sensitive information, then applies configurable replacement, masking or removal.
For data and engineering teams reviewing names and contact details before exporting tickets, documents or shared tables.
The MIT-licensed code is free. Compute, hosting and external recognizers can cost extra.
Inspect entity types and matched text, then review potential misses and false positives.
Apply replacement rules to selected entities and compare the text before and after.
Automated detection can miss or mislabel items. Keep human-labeled samples for your languages and file types, and review important output.
A local deployment can use local models and rules. Check what data is sent if you configure an external recognizer.
The official Streamlit tutorial screenshot predates the transition. Data Privacy Stack now maintains the community project.
The MIT-licensed code is free. Compute, hosting and external recognizers can cost extra.
Flag likely sensitive fields in text and mask or replace them with configured operators.
Visit website ↗Submit once. Get listed permanently after approval.
