What agencies ask before a pilot

No. FMIR is a private technology company. Regulatory authority sits entirely with the agencies. FMIR supplies the software those agencies use to carry out supervision they are already mandated to perform — it makes no determinations of its own and holds no enforcement power.

No, and we advise against it. The platform is designed to start with a single agency and a single module — most commonly report triage at the financial intelligence unit — and widen once that is operating in production. Cross-agency features activate as other agencies join.

Screening draws on published sanctions and watchlist data, open blockchain data, and licensed commercial threat intelligence. Sources are disclosed per assessment so an analyst can weigh how a score was produced rather than accept it at face value.

No. Reporting entities interact only with the filing portal — submitting reports and attestations and viewing their own submission history. Risk scores, cross-agency case records and enforcement views are visible to agency users only.

In-country, under the retention and access requirements set by the contracting agency. Deployment can be dedicated cloud or on-premises where required. Access is role-scoped and every read of a case record is logged.

A scoped engagement on one module with a defined report population, typically over one quarter, measured against the agency’s current handling time and escalation accuracy. Pilots run on the agency’s own data under a data-processing agreement.

No. Scoring orders the queue; it does not close cases. Every confirm, downgrade and escalation is made by a named analyst and recorded with a rationale. Automatic clearance applies only to routine threshold filings, and that threshold is set by the agency.

The agency’s data is exported in full in an open format and deleted from our systems on the schedule set in the contract. We do not retain agency case data for model training or any other purpose.