Stop Fraud
Before It's Paid
ClaimGuard scores every claim submitted to your scheme in real time — surfacing fraud patterns across all your provider hospitals that no single facility could ever see alone.
See It In Action
Trusted by leading insurance schemes to safeguard claims integrity across hundreds of hospitals and millions in payouts.
Fraud Detection Across the Board
Every claim is scored against six families of fraud signals — combining facility-level history with patterns only visible at scheme scale.
Amount Anomalies
Claims priced abnormally high for the diagnosis or service mix, benchmarked against national medians and peer facilities.
Duplicate & Near-Duplicate Claims
The same patient and treatment claimed more than once, even when invoice numbers and timestamps are deliberately varied.
Cross-Facility Patterns
The same patient claiming treatment at multiple hospitals in a short window — visible only with scheme-wide data.
Service–Diagnosis Mismatches
Billed procedures and medications that are not clinically indicated for the submitted diagnosis code.
Provider Outliers
Facilities whose claim volume, average value, or service mix spikes suddenly against their own baseline.
Timing Anomalies
Claims submitted in suspicious batches or at unusual hours — a known signal of automated and coordinated abuse.
How it works
ClaimGuard slots into your existing OpenIMIS instance via the documented module signal pipeline. No rip-and-replace, no parallel system.
Claim arrives
A hospital submits a claim into OpenIMIS — exactly as today, no workflow change.
AI scores in real time
ClaimGuard scores it against six fraud families and scheme-wide patterns.
Reviewer decides
High-risk claims surface in the queue with plain-language reasoning. Human-in-the-loop.
Payment protected
Clean claims pay through. Suspect claims are held, investigated, or rejected with evidence.
Why Choose ClaimGuard
We give insurance schemes the cross-hospital visibility no single facility can have — and slot directly into the infrastructure you already run.
Cross-Hospital Visibility
Detect patient and provider patterns no individual hospital can see — the signal lives in the scheme, not the facility.
Built on OpenIMIS
Integrates directly into your existing claims infrastructure. No rip-and-replace, no parallel system to maintain.
Trusted by schemes that pay at scale
"Within the first month we caught a cross-facility ring that had been billing the same patient at four hospitals in two weeks. No legacy tool flagged it."
The team behind ClaimGuard
A small, focused team of engineers, fraud analysts and public health specialists building for African health financing.
Built for every scheme size
Start with a pilot. Scale when the evidence speaks for itself.
90-day pilot on a single OpenIMIS instance.
- Up to 5,000 claims/mo
- All six fraud families
- Email support
Production deployment for national or regional schemes.
- Unlimited claims
- Cross-facility analytics
- SLA & dedicated reviewer training
- Audit-ready reporting
For consortia rolling out OpenIMIS across multiple payers.
- Multi-tenant deployment
- Custom model tuning
- On-site enablement
Common questions
OpenIMIS catches exact duplicates. ClaimGuard catches statistical anomalies, near-duplicates, cross-facility patterns, service–diagnosis mismatches, and provider outliers — patterns that pass the existing checks.
No. ClaimGuard is human-in-the-loop. The AI scores and explains; a reviewer decides. High-risk claims are surfaced with evidence, never silently denied.
No. ClaimGuard installs as a standard OpenIMIS backend module via openimis.json. It listens to the documented claim submission signal — no fork, no core changes.
We tune precision per scheme and surface reasoning with every flag, so reviewers can dismiss false positives in seconds. The system learns from your decisions.
Inside your scheme's environment. ClaimGuard runs against your OpenIMIS database; only de-identified signals leave your perimeter when you opt into benchmarking.
Ready to protect your scheme?
Book a 30-minute walkthrough on real claim data. We'll show you what your existing pipeline is missing — live.