Data and methodology
Trust infrastructure starts with data provenance.
Compliance buyers care about provenance. This page explains where data comes from, how status is labelled, and how Thesmios avoids pretending that every match is a verified fact.
Sources
Corporate
FreeCompanies House, Charity Commission, Insolvency Service, London Gazette
Regulatory
FreeFCA Register, SRA, GMC, NMC, HMRC MLR
Sanctions
FreeOFSI, OFAC, EU, UN, OpenSanctions
Courts and tribunals
FreeBAILII, Employment Tribunal judgments, civil records where available
Media and public web
FreeGoogle News, Bing News, public social content where consented
Premium planned
PlannedComplyAdvantage, OpenCorporates, Land Registry, LexisNexis
Verification levels
Verified and self-declared data must never look the same.
Source-verified
Checked against an authoritative register, API, issuer, or provider.
Self-declared
Provided by the subject, labelled clearly, and optionally escalated for review.
Public-source match
Matched from public records with match strength, source link, and review status.
Entity resolution
False positives are a product problem, not a footnote.
Phase 1 uses deterministic matching and search-based entity resolution with reviewer-visible match factors. Phase 2 adds specialist entity resolution tooling and stronger deduplication.
The target is not to hide uncertainty. The target is to show match strength, source quality, review status, and whether the subject has confirmed, disputed, or contextualised the finding.
Refresh cadence
Static credentials
Refreshed on expiry, change, or user update.
Public lists
Refreshed according to source cadence and risk tier.
Monitoring
Daily or event-driven where supported by source and plan.