Folio UK is designed to be inspectable from end to end. This page sets out our retrieval, provenance, verification, entity resolution and audit methodology in plain English.
Folio UK fetches data only from primary, official sources — UK Parliament APIs, Companies House, Contracts Finder, Find a Tender, the Register of Consultant Lobbyists and gov.uk transparency publications. No third-party intermediaries. Retrievals are timestamped and stored alongside an immutable snapshot of the source response.
Every record carries a complete provenance envelope: source name, source URL, source identifier, retrieval timestamp, verification timestamp, verification status and snapshot identifier. Users may inspect, compare and verify any record directly against its primary source.
Records are passed through deterministic verification checks before being marked verified: schema validity, source-identifier integrity, cross-source corroboration where applicable. Records that fail any check are marked pending or routed to a human review queue.
Entities are resolved deterministically using stable source identifiers — never by AI matching. Possible matches that do not meet a confidence threshold are sent to a manual review queue. Resolution decisions are versioned and reversible.
Every change to a canonical record produces an audit log entry — previous value, new value, change summary, snapshot reference. Audit trails are immutable.
Folio UK does not infer wrongdoing. Folio UK does not score, rank, accuse or speculate. Folio UK does not generate AI conclusions about people, institutions or events. The platform surfaces the public record.