Presidio

Microsoft built Presidio as an open-source SDK for detecting and de-identifying personally identifiable information in text, images, and structured data, released under the MIT license. The Analyzer combines named entity recognition, regular expressions, rule-based logic, and checksum validation to find entities such as credit card numbers, names, locations, phone numbers, and national IDs across multiple languages, with pluggable NLP engines and support for custom recognizers and external models. The Anonymizer then redacts, masks, hashes, or replaces detected values, an Image Redactor removes PII text from images including DICOM medical files, and a structured data module covers tabular sources. It deploys as Python packages, Docker containers, on Kubernetes, or inside PySpark pipelines. The documentation is candid that automated detection cannot guarantee catching every sensitive value, positioning Presidio as one layer in a privacy program. Typical users are data engineering and compliance teams adding de-identification to pipelines and LLM applications.