Detectory
All industries

Flagship

Enrollment fraud detection for state-based marketplaces

Our home turf. Detectory was purpose-built for the enrollment front door of state-based exchanges: SEP proof forensics, outstanding-verification queue automation, broker-ring detection, and referral-grade audit trails that survive CMS, your MFCU, and an oversight hearing.

The problem

The GAO walked 23 fake applications through the front door

In December 2025, GAO investigators submitted 24 fabricated applications to marketplaces using fake documents and invented identities. 23 were approved. Meanwhile, unauthorized broker enrollments generated over 200,000 consumer complaints to CMS in a single year, and improper payments across health programs run to $186B annually. Marketplace eligibility teams are staring at uploaded paystubs and loss-of-coverage letters with no forensic tooling, deciding federal subsidy eligibility by eye.

23 of 24

GAO fake applications approved

GAO, Dec 2025

200K+

unauthorized-broker complaints to CMS

CMS, 2024

$186B

annual improper payments, health programs

GAO estimate

Example verdicts

The documents your team sees, scored

Representative documents from this workflow with the forensic verdict Detectory renders and the finding behind it. Suspect documents are never auto-denied; they route to your reviewers with the evidence itemized.

Loss-of-coverage letter (SEP)

Employer letterhead, uploaded to claim special enrollment

Likely fraudulent

Finding: Identical letter body reused across 31 applications from 3 broker NPNs; employer EIN does not resolve

Paystub (income OV)

ADP-format stub for outstanding income verification

Likely fraudulent

Finding: YTD totals do not sum; PDF authored in a design tool, not a payroll system

Marriage certificate (SEP)

County-issued certificate for marriage SEP

Genuine

Finding: Seal geometry, registrar signature, and county template all consistent; cleared in 11 seconds

I-766 work permit (immigration OV)

Employment authorization card for status verification

Human review

Finding: Card format valid but USCIS number issued in a different name per cross-application history; routed to investigator

What Detectory does here

Purpose-built detection for this workflow

SEP proof forensics

Every special-enrollment document is scored for tampering, template reuse, and issuer validity, targeting the exact channel GAO exploited.

OV queue automation

Genuine documents clear in seconds so legitimate enrollees beat the 90-day deadline. Suspect uploads route to humans with itemized findings.

Broker-ring analytics

Velocity, shared-template, and device-cluster signals across broker NPNs surface organized rings, not just single bad applications.

Cross-program duplicate detection

The same identity enrolled in both the marketplace and Medicaid is flagged so federal funds are never paid twice.

Case walkthrough

A broker ring hits your SEP channel

Over 48 hours, one broker NPN submits 22 loss-of-coverage SEP applications. Each letter looks plausible alone. Detectory clusters them: identical letter body with swapped names, the same PDF creation tool, and three device fingerprints across two more affiliated NPNs. Every application is held for human review before any subsidy flows.

Outcome: Your integrity team gets one consolidated case with 22 itemized exhibits, and a one-click referral packet formatted for your MFCU and CMS. No legitimate enrollee was touched.

See Detectory score your own redacted SEP and OV documents in a 30-minute walkthrough.

Other industries