CamoText was born from a practicing lawyer's experience with addressing client confidentiality and data privacy while using LLMs for text analysis and form generation: manual redaction was too time-consuming and existing cloud- and API-based solutions posed extensive privacy and compliance issues. CamoText is a tool for users to anonymize text fully offline and locally on their machine, detecting and redacting PII and other sensitive text before any external transmission. It uses bundled NLP and custom pattern-matching to find and obscure PII and numerous other categories of private data including:
- names of individuals and entities,
- addresses,
- account numbers,
- contact details,
- monetary amounts and financial information,
- URLs and IP addresses,
- filenames and locations,
- UUIDs, and more.
It supports intuitive human-in-the-loop review of the output: a straightforward highlight-to-anonymize feature enables quick redaction of any additional text deemed sensitive. Zero data is retained when the application is closed.
CamoText has a simple user interface, with minimal technical requirements met by standard Windows or Mac laptops. Tailored enterprise or public sector packages are available, with custom data recognizer options, whitelabeling, and user training for using the most powerful general purpose LLMs in a compliant manner.
Loading...