Honest-ML

An open source Python framework for diagnosing and mitigating bias in machine learning models through automated fairness auditing.

Honest-ML provides a streamlined toolkit for data scientists to quantify model equity. It integrates directly into Scikit-learn workflows to track disparate impact across demographic slices (like age, gender, or race) using metrics such as equalized odds and demographic parity. By generating clear visualizations and bias scores, the library enables teams to identify algorithmic drift and implement re-weighting strategies before deployment. It is built for production environments where transparency and regulatory compliance are non-negotiable.

https://github.com/honest-ml/honest-ml