Skip to content

For health plans

Fraud and care models fail on provider and eligibility data first.

Prior authorization, payment integrity, provider data. Health plans are funding AI for all three, and most of it fails on provider identity, adjusted claim lines and retroactive eligibility. Before your vendor goes live, we profile the systems the use case depends on, grade each one, and give you a written verdict: go, fix then go, or no-go.

Already working with a prior authorization or payment integrity vendor? We grade your data before it goes live.

  • 6 weeks from data access
  • From $75K, fixed fee
  • Profiled, not surveyed
AICore Claims (837)B-Eligibility (834)BProvider DirectoryDUM / Prior AuthC-Pharmacy FeedCEDWB-Prior authorization modelLAKESHORE COMMUNITY HEALTH PLAN (FICTIONAL) · OVERALL GRADES

How It Fails

Health plans are funding AI. Most pilots fail on the data. Here's how it happens.

The pilot below is illustrative.

Prior authorization model · accuracy · illustrative

PRODUCTION DATA100%75%50%89%Week 8The demo wins.Week 12Production data arrives.61%Week 19Accuracy collapses.Week 26Quietly shelved.

Week 8

The demo wins.

Trained on closed authorization cases, with provider and member records cleaned by the utilization management team, the prior authorization model scores 89% in the steering committee demo. Funding approved.

Week 12

Production data arrives.

Live requests start flowing: the same clinician under several NPIs and taxonomy codes, directory records that don't match the requesting provider, clinical rationale in faxed PDFs.

Week 19

Accuracy collapses.

The model can't tell which provider is asking, or whether they're in network. Accuracy slides to 61%, and nurses go back to reviewing every request by hand.

Week 26

Quietly shelved.

Nobody agrees why it failed, so the next budget cycle funds a different vendor. The same data is waiting for them.

Why: The Verdict finds the gap before the build starts, not after go-live.

Use Cases We Grade Against

Every grade is weighted to the use case you're delivering.

Why: A D on a system your use case never touches can wait. An F on the one it depends on can't.

Data QualityDocumentationLineageAccess & SecurityAI ConsumabilityDECIDINGSUPPORTING

What We Grade

The systems usually in scope.

Scoped by 5, 10 or 15 source systems: the ones your use case reads from, not the whole estate.

ClaimsProvider dataEligibility & enrollmentUtilization managementData warehouseClinical documentsYour AIuse case

Claims

Lines as originally billed, and every adjustment since

Provider data

One identity per clinician across NPIs and TINs

Eligibility & enrollment

Coverage as of the date of service, not today

Utilization management

Authorization decisions and their rationale

Data warehouse

Where lineage is usually lost

Clinical documents

PHI a model needs to read, under tight access

Sample report carda fictional health plan
System
Core Claims (837)BCCBC
Eligibility (834)BBCBB
Provider DirectoryFDFCD
UM / Prior AuthCDDCD
Pharmacy FeedCCCBC
EDWBCCBC

This is where a prior authorization pilot dies.

PHI stays where it is.

Profiling runs inside your environment by default, with minimum-necessary access. We sign your NDA and complete your vendor security review before any data is touched.

Know before you fund the pilot.

A Data Readiness Verdict in six weeks: go, fix then go, or no-go, from your real data. From $75K for up to five source systems, set in the scope before we start.

See exactly what's included in the Verdict