ML Audit Library

Solves Data Lineage Blindness by tracking granular preprocessing steps.

ml-audit is a lightweight Python library designed to bring transparency and reproducibility to data preprocessing. Unlike standard experiment trackers that treat preprocessing as a black box, this library records every granular transformation applied to your pandas DataFrame.

📌Key Features

• Full Audit Trail: Automatically logs every step (Imputation, Scaling, Encoding, etc.) into a JSON audit file.

• Reproducibility: Verify if your data pipeline produces the exact same result every time using hash validation.

• Visualization: Auto-generates an interactive HTML timeline of your preprocessing steps.

View On Github View on PyPI View Documentation