Medicare Claim Attestation / Architecturejilsovereign.com/products/medicare-claim-attestation/architecture
Architecture

Medicare Claim Attestation Architecture

DRG upcoding, phantom billing, credential fraud, Part D abuse

Purpose-built attestation for Medicare and Medicaid claims. DRG upcoding, phantom billing, credential fraud, and Part D prescription abuse detected before settlement or identified in historical claims, with court-ready evidence on every finding. Powered by 28 federal public data feeds and the full Verdict Engine.

28
Federal data feeds
148
Verdict Engine checks
RADV
Defense-grade
FCA
60-Day Rule timely

§ 01Business Architecture

AudienceMCOs, state Medicaid integrity programs, RAC contractors, qui-tam firms post-CREB™.
BuyerGC, CCO, VP Payment Integrity, RAC team lead.
Economic ModelPer-claim flat fee for pre-settlement; per-batch retro pricing; RADV defense subscription tier.
PartnersPlaid (banking fingerprint), CMS PECOS direct connection, OIG/State exclusion feeds.

§ 02Technical Architecture

Claim parser

EDI 837/835, FHIR R4, custom CSV ingestion.

28-feed manifest

NCD/LCD/MUE/NCCI/Pub 100/PECOS/NPPES/LEIE/SAM/Open Payments/Provider Enrollment/etc.

Verdict Engine

Medicare-relevant subset of 175 checks - DRG, phantom-billing, credential, Part D, dual-eligible.

Pattern matcher

Cohort-baseline pattern engine - DRG-shift, length-of-stay outliers, billing-volume anomalies vs peers.

CourtChain™ seal

Each finding sealed with source citation + cohort baseline + finding pattern.

Data SourcesCMS NCD/LCD/MUE/NCCI/Pub 100/PECOS/NPPES, OIG LEIE, SAM.gov, Open Payments, ProPublica/Public Census, plan medical policy.

§ 03Process Flow

01
Claim ingested
Pre-settlement: live. Retroactive: batch.
02
148-check fan-out
Medicare-relevant subset runs deterministically.
03
Cohort baseline
Each finding cross-referenced against PEPPER + Medicare population baselines.
04
Verdict + seal
Hit/Clean/Conditional + sealed CREB™.
05
RADV + qui-tam
Bundle pre-built for RAC + FCA proceedings.

§ 04Plain English Example

Worked example

Plain English: a $850K DRG cohort flagged for upcoding

MCO ships Q4 inpatient batch (3,200 cases, $850K). Engine fires: 47 DRG-shift findings (DRG 470 to 469 in cohorts where the case-mix index doesn't support it; PEPPER baseline shows the hospital is in the 99th percentile for this shift). 12 false-positives. 35 confirmed upcoding pattern. Estimated overpayment: $164K. CREB™ sealed for each. MCO holds payment, refers to RAC. Within 27 days (FCA 60-Day Rule window), MCO has filed self-disclosure with CMS, avoiding qui-tam exposure.

§ 05Capabilities Summary

DetectionDRG-shift, phantom-billing, credential-fraud, Part D abuse, dual-eligible misclass
Cohort baselinePEPPER + Medicare population - finding contextualized to peer hospital cohort
ModesPre-settlement (live) + retroactive (batch)
DefenseRADV + FCA 60-Day Rule timeline, sealed bundle ready
Throughput<2s P95 pre-settlement; <12 min/M claims retroactive