Provenance

Provenance is the documented trail describing where a specific record or dataset actually originated, what license or public-record basis supports its inclusion in a corpus, and what verification, including deduplication, it has undergone since being ingested. Provenance documentation is what separates a checkable corpus claim from an unverifiable one: a total record count presented without source-by-source provenance behind it is a figure a diligence team cannot actually confirm, and an institution presenting only the total, without the underlying provenance, is asking a counterparty to trust the figure rather than verify it. Proper provenance documentation becomes especially important for evidence classes with more complex origins than a straightforward public court filing, medical lien data fragmented across providers and negotiators, or datasets acquired through a merger or a prior initiative whose original assembly history may be incomplete by the time they reach a new custodian. An institutional data room built for a model-driven outcomes platform should document provenance source by source, not present a single headline figure without the supporting trail behind it, since the trail, not the figure, is what a diligence team can actually independently confirm, and a figure presented without it is, functionally, unverifiable regardless of how large or precise it looks.