CG
ClaimGuardNG
Live Prototype Available — Built for NHIA, HMOs & Providers

Nigeria's health claims move through five separate systems. This is the one that watches all of them.

ClaimGuard NG verifies, adjudicates, and pays health insurance claims from any provider — hospital, clinic, pharmacy, lab, or imaging centre — for any HMO, in one national system. An AI engine checks every claim against 15 fraud and integrity rules before a human ever sees it. A second, fully independent engine watches the clinical data underneath for hypertension, diabetes, pregnancy risk, and medication non-adherence — before they become expensive emergencies.

● 5 provider types supported natively ● 15 fraud & integrity rules ● Built on Doktorconnect health infrastructure
CLAIM TRACE PROCESSING
Watch a real claim move through the system — click below to try another
St. Augustine Diagnostic Lab
CLM-2026-0048213 · Lagos · Outpatient Lab
₦18,500
FBC + Lipid Panel
Submitted
AI Verified
Tariff Check
HMO Review
Settled
Time in system: 00:00
5 Types
Provider categories — hospital to pharmacy
2 Engines
Claims AI + Predictive Risk, fully separated
4 Portals
Provider · HMO · NHIA · Enrollee
15 Rules
Fraud & integrity checks, 14 live today
Any HMO
One platform, every health plan administrator
How It Works

Two engines. Deliberately kept apart.

Claims adjudication and health risk prediction answer different questions, for different audiences. ClaimGuard NG never lets one engine's output quietly become the other's input — whether a claim gets paid is decided on claims data alone.

Engine A · Runs on billing data
Claims AI
  • 01
    Claims VerificationConfirms eligibility, provider accreditation, and plan match before a claim enters review.
  • 02
    Fraud & Integrity DetectionChecks 15 distinct rule types — duplication, coding, thresholds, behaviour, and more.
  • 03
    Billing ValidationChecks every line item against the current NHIA tariff schedule.
  • 04
    Claims ReconciliationMatches paid claims against HMO disbursement records — closes the loop on every naira.
Engine B · Runs on clinical data
Predictive Risk
  • 01
    Hypertension RiskTracks BP readings and refill continuity to flag members drifting out of control.
  • 02
    Diabetes RiskMonitors HbA1c trends and prescription continuity for members with diabetes.
  • 03
    Pregnancy ComplicationsWatches antenatal visit cadence and risk factors through every trimester.
  • 04
    Medication AdherenceDetects refill gaps against expected dosing schedules.
  • 05
    Preventive IndicatorsSurfaces members overdue for screenings or follow-up.
Who It's For

The same system, three different reasons to trust it

A regulator, an HMO claims officer, and a hospital records team are not asking the same question. Pick yours.

Right now, no one sees every claim across every HMO at once.

ClaimGuard NG gives NHIA a single oversight dashboard across every HMO, every provider, and every state — fraud heatmaps, HMO scorecards, and claims velocity, updated as claims move, not in a quarterly report.

  • Fraud flags visible across HMOs, not siloed inside each one
  • HMO performance scorecards — flag rate, resolution speed, claims value
  • Population health signals from Engine B, aggregated for policy use
  • A published rule registry — every flag is auditable, not a black box
What this replaces
Fraud caught months later, in a post-payment audit, after the money has already moved.

Your claims officers shouldn't have to re-check what a machine already checked.

Every claim arrives in your queue with eligibility, tariff conformance, and 14 live fraud rules already run — sorted by risk, not by date received, with the evidence trail attached so a decision takes seconds, not a phone call.

  • Clean claims fast-tracked automatically — most claims need no review at all
  • Flagged claims arrive with the exact rule, the exact number, and why it fired
  • One-click Approve, Query, or Reject — decision and reasoning logged together
  • Works alongside your existing core system — claim and adjudication exchange, not a rip-and-replace
What this replaces
A claims officer manually re-pricing every line item against a tariff sheet, all day, every day.

A clean claim should be paid in days, not chased for months.

The claim form adapts to what you actually do — a hospital admission isn't asked to fit a pharmacy's shape. Submit, and watch your own claim move through verification and validation in real time, the same trace an HMO officer sees.

  • Five claim types, each shaped for its provider — bundled, fee-for-service, dispensing, panel, procedure
  • Real-time eligibility check before you even submit
  • See exactly why a claim was flagged — no guessing, no resubmission roulette
  • Payment and remittance tracking in one place
What this replaces
Submitting a claim and finding out six weeks later, with no explanation, that it was rejected.
Fraud & Integrity Detection

15 rules. Six categories. Nothing hidden.

Every rule the Claims AI can raise is published — the same registry the live engine evaluates against. A sample of what it watches for:

Coding Integrity
Upcoding Pattern
A higher-acuity procedure code billed than the encounter documentation supports.
Duplication & Existence
Phantom Encounter
A billed service with no matching facility check-in record — the encounter can't be confirmed.
Threshold & Policy Limits
Policy Limit Proximity
A claim that brings an enrollee within range of their annual benefit category limit.
Utilization Pattern
Length-of-Stay Outlier
An inpatient stay well outside the expected range for its diagnosis-related group.
Provider / Patient Behaviour
Provider-Network Anomaly
A patient billed by an unusually high number of unrelated providers in a short window.
Logical Consistency
Geographic Implausibility
The same patient billed at two facilities too far apart for same-day attendance at both.
Interoperability

This is a clearing house, not just a portal.

A hospital's EMR, a pharmacy's POS system, or an HMO's core platform shouldn't have to learn ClaimGuard NG's internal data model to talk to it. They speak the formats they already speak — ClaimGuard NG translates.

Comes In As
X12 837 (EDI)
HL7 FHIR R4 Claim
Manual Portal Entry
Becomes
One Canonical Claim Object
Runs Through
The Same Claims AI Engine
No second set of fraud rules for API-submitted claims
● LIVE837
Claims Submission
Institutional & professional claims via X12 EDI — the format most hospital EMRs and HMO core systems already export.
● LIVEFHIR
FHIR R4 Claim Resource
The modern REST/JSON alternative to X12 — for newer systems built FHIR-native from the start.
PLANNED835
Payment Remittance
Electronic remittance advice sent back to the provider automatically once a claim is adjudicated.
PLANNED270/271
Eligibility Inquiry
Real-time eligibility check before a claim is even submitted — fewer claims rejected for coverage reasons.
PLANNED276/277
Claim Status Inquiry
A provider's EMR polls claim status directly — no one has to log into the portal to check.
PLANNEDFHIR Encounter
Facility Check-In Confirmation
Auto-confirms an inpatient encounter actually happened — closes the gap the Phantom Encounter rule watches for today.
Built For Every Provider Type

One platform. Five claim shapes.

Hospital
Bundled / DRG claim
Outpatient Clinic
Fee-for-service
Pharmacy
Dispensing claim
Diagnostic Lab
Test-panel claim
Imaging Centre
Procedure claim
Four Portals

One claim. Four points of view.

P
Provider
Submission Portal
Submit and track claims. The form adapts automatically to provider type.
H
HMO
Adjudication Portal
Review what the AI already checked, decide what only a human should.
N
NHIA
Oversight Dashboard
Every HMO, every provider, every claim — one regulatory line of sight.
E
Enrollee
My Claims
See exactly what was billed in your name. Dispute it in one tap.
Questions

Before you ask

It's a working prototype — submit a real claim through the Provider Portal and watch it move through the same AI engine an HMO officer would see, with the same fraud rules actually evaluating it. Nothing on the live trace above is staged; it runs the real verification logic.
No, and that's deliberate. Engine A (claims) and Engine B (clinical risk) never share inputs. A flagged health risk is visible to the HMO and to the enrollee as a population-health signal — it has no path into the adjudication decision.
No. ClaimGuard NG is designed to exchange claims and adjudication results with an HMO's core system via API or batch file — it sits alongside, the same way it can draw eligibility data from the NHIA Enrollee App or encounter data from a provider's FHIR R4 records.
Hospitals (bundled DRG claims), outpatient clinics (fee-for-service), pharmacies (dispensing claims), diagnostic labs (test-panel claims), and imaging centres (procedure claims) — each with its own claim shape, not a generic form forced onto five different realities.
14 of 15. The one exception — Prescription Pattern Anomaly — needs months of real refill history to detect reliably, which a prototype can't fabricate honestly. It's published in the rule registry, ready to activate once real prescription data accumulates.

Stop reading about it. Submit a claim and watch what happens.