AI recipe
Training Proof
Dataset hash + hyperparameters. Signed.
What it does
Hashes the training dataset and hyperparameter configuration, then signs the bundle with the training run's timestamp. Proves what data was used and when — useful for copyright claims, regulatory audits, and replication disputes.
Who this is for
ML engineers and research labs who need a durable record of training provenance for models that may be commercially deployed or legally challenged.
Lives in
This recipe belongs to the AI domain — Every model, every decision — governed and proven.