Feature Provenance Ledger

Feature Provenance Ledger is a ml record that tracks where data came from and how it changed for input signals used by a machine learning model. It uses hashes, source labels, and transformation history so teams can audit model inputs reliably while keeping evidence, reliability, and public-safe operational boundaries clear.

“The machine learning team used Feature Provenance Ledger when a feature distribution shifted, so the team could audit model inputs reliably before the model moved into evaluation.”