🛠 Incognito is under active development

️ Keep an eye out on GitHub for releases

Introducing Incognito

In the digital age, protecting personal data is paramount. The GDPR outlines that data is considered anonymous when it cannot be linked back to an identifiable person. Achieving this in open text data presents unique challenges, given the myriad of attributes that can lead to personal identification. Incognito is designed to tackle this head-on by obfuscating sensitive information such as names, contact details, and locations, which are closely tied to individual identities.

Our approach with Incognito is pragmatic and pattern-based, recognizing that while the need for anonymization is universal, the specifics can often be addressed through recognizing and mitigating common data attributes.

How Incognito Works

Incognito employs advanced named entity recognition combined with rule-based logic to pinpoint and anonymize personally identifiable information (PII) within text data. Our models are meticulously trained to ensure fairness and reduce bias across gender, ethnicity, and other personal identifiers, making Incognito a robust tool for privacy preservation in diverse datasets.

FAQ

Does Incognito guarantee anonymity?

While absolute anonymity can be challenging to guarantee due to the nature of data, Incognito significantly reduces the risk of identification by removing or obfuscating common identifiers. This approach is typically sufficient for most anonymization needs.

Can Incognito be self-hosted?

As an open-source solution, Incognito supports self-hosting, allowing organizations to integrate it within their own data processing pipelines in compliance with its permissive licensing terms.

Incognito is under active development, and we're seeking feedback and collaborations to refine its data privacy capabilities. If data privacy and ethical AI are your passions, reach out to explore how you can contribute to Incognito's journey.