Developers

One API call for sanctions, politically exposed persons (PEPs) and adverse media

POST a subject and the feeds to run. Minerva returns ranked potential matches, each with its score, the risk flags that fired and the sources behind them. Entity matching reads every hit in context, so your review queue takes up to 75% fewer false positives.

Book a demo Browse Minerva docs

Screening response3 potential matches

1.2 s

John Smith

Individual · born 1984 · Canada

Clear

0.11score

no matchcriteria match

No list returned a hit above threshold.

Mary Roe

Individual · born 1971 · Cyprus

Sanctions match

0.94score

closecriteria match

OFAC SDN List · name and country context agree.

Acme Inc

Entity · incorporated 2014 · Singapore

Needs a person

0.42score

partialcriteria match

PEP register · one officer in common, no identifier match.

300ms to 3s: typical sanctions and PEP search
Up to 75%: fewer false positives
~5 days: typical implementation

Coverage

Screen and onboard up to 96% of clients without a manual review

Three feeds, one request. You pick which of them a given search runs, per call or per workspace, and Minerva resolves the results into one entity before it scores anything.

Book a demo More on Solutions

Sanctions
Screen against sanctions lists from 197 list-issuing jurisdictions, refreshed daily and matched using identity context.
Sanctions lists
From 197 jurisdictions
List refresh
Daily
Matching
Context-aware entity resolution
Scope197 counts list-issuing jurisdictions, which include supranational and territorial issuers, rather than UN member states.
Politically exposed person (PEP) identification
Lists are the floor. Minerva also reads open-source media, then grades the exposure so your policy decides what is worth an alert.
Coverage
Global
Beyond lists
Open-source media
Exposure tiers
Configurable, 1 to 4
Adverse media and open-source intelligence (OSINT)
Real-time web search with sentiment analysis, so a name that appeared in this morning's reporting is not waiting for a list refresh.
Media sources
300,000+
Languages
150
Full search
Within 45 seconds
Triage
Sentiment, to your criteria
Scope45 seconds is the ceiling for a full enhanced due diligence search across open sources. A typical sanctions and PEP search returns in 300ms to 3s.

The request

See the payload before you book the call.

Screening is synchronous: the 200 carries the result itself, not a job id to poll, so the call can sit in your onboarding path rather than behind a queue. It returns up to ten ranked potential matches, each with the source lineage you will need when an examiner asks where a value came from.

Browse Minerva docs

Request · POST /v1/search-sync

01POST /v1/search-sync HTTP/1.1
02Host: api.gominerva.com
03Content-Type: application/json
04x-api-key: $MINERVA_API_KEY
05 
06{
07  "type": "individual",
08  "name": "John Example",
09  "year": 1970,
10  "country": "Canada",
11  "feeds": ["Sanctions", "PEP", "News"]
12}

One POST, and the 200 carries the screening

The call is authenticated with an x-api-key header, and the body that comes back is the screening itself rather than a receipt for one. type echoes the request's subject as its numeric code. Keep the searchId: passed as searchResultId to POST /v1/reports, it generates the PDF audit export for this search.

Up to ten ranked potential matches

results holds the ranked potential matches. Each one is a profile, and the name block quotes the value that matched with its own match_score and criteria_match_level.

The number a review workflow branches on

score is the overall Criteria Match Score from 0 to 1. The API reference describes 1.0 as exact, 0.80 to 0.99 as partial, and scores below 0.80 as low confidence. Field-level values such as name.criteria_match_level use the labels exact, close, loose, or none.

Which feeds fired

checklist.screen answers the request's own feeds array: Sanctions, PEP, News and the rest, each one true or false for this match.

The decision and its lineage

review_status is the state your workflow reads. In Hint Mode, disposition_hint adds the advisory prediction, confidence and rationale while sources carries the lineage behind the profile.

Response · 200 OK · trimmed

01{
02  "status": "success",
03  "code": 200,
04  "duration": 1.2,
05  "searchId": "b426f9b4-17ed-4f31-a39d-c49f45c753f0",
06  "type": 0,
07  "page": 1,
08  "pageSize": 10,
09  "count": 1,
10  "results": [
11    {
12      "name": {
13        "value": "John Example",
14        "match_score": 0.97,
15        "criteria_match_level": "close"
16      },
17      "score": 0.97,
18      "checklist": {
19        "screen": { "Sanctions": true, "PEP": false, "News": false }
20      },
21      "review_status": "unresolved",
22      "disposition_hint": {
23        "prediction": "true_positive",
24        "confidence": 0.99,
25        "analysis_status": "completed",
26        "rationale": "The searched subject matches the potential match on name and country context, and no discovered identity attribute contradicts the match."
27      },
28      "sources": ["OFAC SDN List"]
29    }
30  ]
31}

What your integration reads

The fields most likely to become branches, records or analyst context in your integration.

Key screening API response fields
Field	Type	What it carries	Permalink
`searchId`	string	Pass it as searchResultId to POST /v1/reports to generate the PDF audit export for this screening.	#
`type`	number	Subject type. The request takes the word (individual) and the response returns the numeric code (0) for that same subject.	#
`score`	number	The overall Criteria Match Score from 0 to 1. The API reference describes 1.0 as exact, 0.80 to 0.99 as partial, and scores below 0.80 as low confidence. Field-level values such as name.criteria_match_level use exact, close, loose, or none.	#
`checklist.screen`	object	Which risk feeds fired for this match: Sanctions, PEP, News and the rest, each true or false.	#
`review_status`	string	not_found, unresolved, true_positive, false_positive or suppressed. New matches start unresolved.	#
`automatic_disposition`	object	The prediction applied in Full Auto Mode when a match carries a risk flag.	#
`disposition_hint`	object	The advisory prediction in Hint Mode, with its confidence, analysis_status and rationale.	#
`sources`	string[]	Every source that contributed data to the profile, which is the lineage an examiner asks for.	#

Source: Minerva Core API reference. Payloads are trimmed from the published examples.

What you get

Screening that runs inside your onboarding request

Minerva matches structured and unstructured data to your customer in context rather than on name similarity, so your team is not making those comparisons by hand.

01Lower false positives by up to 75%: Sanctions, PEP identification, adverse media and OSINT run in one request against one resolved entity, so a hit is judged on the whole picture rather than four times on four partial ones. That is where the reduction comes from.
02Onboard the right customers, faster: Automated screening clears up to 96% of clients without a manual review and flags the remaining clients for further review.
03Greater coverage from open sources: Minerva goes past the obvious lists and searches the live web, which is what improves true-positive detection rather than just trimming the queue.
04Data-rich responses with full lineage: The response identifies the sources that contributed to each profile, so your team can retain source lineage with the screening result.

See a screening request with your configuration.

Bring the payload you would actually send. In the demo we run it against your own list configuration and walk the response field by field, including the disposition annotations your integration would branch on.

Book a Demo

A background motif used as part of Minerva's unique branding.