Sanctions Screening

Sanctions screening software, built on entity resolution — not name matching alone

A sanctions match is only as reliable as the entity resolution behind it. Here's how Veredis screens a counterparty, what evidence a match carries, and where a screening result can legitimately come back inconclusive.

The problem

Fuzzy name matching against a sanctions list produces two failure modes at once: it misses genuine matches hidden behind transliteration or abbreviation differences, and it floods reviewers with false positives from common names that happen to string-match a listed entity. Neither failure is acceptable in a regulated workflow — a missed match is a real compliance exposure, and a high false-positive rate trains reviewers to rubber-stamp alerts, which defeats the point of screening in the first place.

Veredis's approach: resolve the entity, then screen it

Before a name is checked against a sanctions list, Veredis attempts to resolve it to a real, identifiable entity — using official registry and ownership data (legal entity identifiers, corporate registry records, and beneficial-ownership signals where available) rather than treating the name string alone as the subject of the check. That resolution step is what lets a screening result distinguish "this is definitely the sanctioned party" from "this name superficially resembles a sanctioned party."

Sanctions-list data itself is versioned: every check records which list version and source was consulted and when, so a screening result from six months ago can be reconstructed exactly as it looked at the time — not re-evaluated silently against today's list and presented as if it were the original result.

Match confidence is shown as an explicit percentage, and direct hits are distinguished from indirect hits on a related party (for example, a parent or affiliated entity) rather than collapsed into one undifferentiated flag.

The evidence model

  • Which sanctions list(s) and versions were checked, and when
  • The entity-resolution path used to identify the counterparty, not just the raw name string submitted
  • Match confidence as a percentage, with the specific matched fields shown
  • Whether a hit is direct (the named party itself) or indirect (an ownership or affiliation relationship)

Where the limitations are

Screening is a point-in-time check against the sources available at the time it runs. If an official source is temporarily unreachable, Veredis reports that check as source-unavailable rather than silently treating it as a clean result — a screening result that comes back "needs review" because a source could not be reached is not the same fact as a confirmed clean counterparty, and Veredis does not blur that distinction.

Veredis's sanctions screening is decision support, not a substitute for a qualified compliance officer's judgment on ambiguous matches, novel sanctions programmes, or jurisdiction-specific nuance a general-purpose rule set was not built to capture.

Governance

See Intelligence Sourcing & Data Boundary for exactly which registry and sanctions signals are served from governed, versioned intelligence versus queried live, and Security for how screening inputs are handled and retained.