InackGuardEvery source, one search.
Screen people and companies against sanctions, exclusion, export-control and forced-labour lists, politically exposed persons, registers, court records and the news, in one pass. Every finding is cited; every gap is named.
What is sanctions screening software?
Sanctions screening software checks a person or company against the lists governments publish of people and organisations you may not deal with, and against exclusion, export-control and forced-labour lists, politically exposed persons and the news. Good screening software searches every source in one pass, scores how close each listed name is to the one you typed, shows the strong and possible matches with their reasons, and records which sources were searched so you can prove it later. InackGuard does all of that and shows the price before each check runs.
Searching one source at a time leaves gaps.
Most teams check a few lists by hand, in different tabs, with different spellings, and save screenshots. Some sources get skipped on busy days, and nobody can show later which ones were searched. InackGuard searches them together, scores every possible match and records exactly what was and was not searched.
What it does
Six capabilities, each one visible in the report it produces.
Fuzzy matching
Listed names are written in many ways: transliterated, reordered, shortened, with a middle initial or without. Matching by exact spelling misses most of them. InackGuard compares tokens, so Sovcomflot Maritime Ltd and SOVCOMFLOT MARITIME score the same, while unrelated names that share one common word stay low.
- Transliterations, aliases and reordered names are scored, not just exact spellings
- Common tokens such as Ltd, Holdings and Trading carry less weight
- A company is not a person: a cross-type overlap is a lead, never a match
RECORDEDThe name as typed, every candidate considered, and the score of each.
Clear tiers
Every candidate lands in one of three tiers: strong, possible or weak. Strong matches go to a person every time. Possible ones go when a detail such as a date of birth or country cannot settle them. Weak overlaps are counted and named in the report, not shown as findings.
- Strong, 85 and above: always reviewed
- Possible, 65 to 84: reviewed unless a detail rules it out
- Weak, below 65: counted, never shown as clear
RECORDEDThe tier of every candidate and the threshold set that produced it.
Country and date of birth
A country or a date of birth is optional, but it does most of the work on a common name. A listed entry with the same date of birth becomes a strong match; one with a different date is set aside, and the report says why. Nothing is guessed when a detail is missing.
- Date of birth agreement raises the score; disagreement lowers it, and the report says so
- Country narrows the candidate set before scoring
- Missing details are recorded as missing, not assumed
RECORDEDThe identifiers given, and which listed identifiers they were compared with.
Named gaps
If a list could not be searched, because a source was down, excluded by your settings or not licensed for your plan, the report names it under not searched. A reader never mistakes a gap for a clear result, and a re-run later picks the gap up.
- Every source has one of three states: searched, not searched, not licensed
- The report's coverage table is part of the sealed record
- Re-screen fills gaps when the source returns
RECORDEDThe coverage table: each source, its state, and the list date used.
People and companies
A person is matched against people, a company against companies, vessels and aircraft against their own entries. The subject type comes from the form or the file, so a shipping company named after its founder is not confused with the founder.
- Subject type set per check or per row in a file
- Entity-type mismatches capped as leads
- Vessels and aircraft screened against their identifiers too
RECORDEDThe subject type and the entity types of every candidate considered.
One name or thousands
A single name in the form, a spreadsheet of five thousand customers, or a call from your own product. Batches run as one job with one price shown first; the API returns the same report shape with an idempotency key so retries never double-charge.
- CSV or Excel up to 5,000 rows per file
- One idempotent API call per subject, signed webhooks when it finishes
- The same sealed record whichever way in
RECORDEDThe way in, the batch or request identifier, and the price agreed before the run.
How it works
- STEP 1Enter a name, upload a file or call the API
- STEP 2Every source is searched together
- STEP 3Matches are scored and tiered with reasons
- STEP 4Review only what needs a person; keep the sealed record
Specifications
| Inputs | Name, optional country, date of birth, identifiers; CSV or Excel for batches |
|---|---|
| Outputs | Result word, ranked findings, coverage table, next steps, sealed record |
| Formats | Web report, PDF, JSON |
| Ways in | App, API with signed webhooks, file upload |
| Speed | Seconds for list checks; minutes for full reports |
Guides that go with it
Written from the product, kept with it.
Questions
How do you reduce false positives?+
Matches are tiered, optional details such as date of birth sharpen the score, companies and people are matched as what they are, and decisions are remembered on re-screening.
Can we screen in our own product?+
Yes, through the API with idempotent requests and signed webhooks.
What happens if a source is down?+
The report names it as not searched. It is never shown as clear.
Reports built on it
See screening on your own names.
Free preview, no card.