Retroactive Payment Audit
Find the money you've already lost.
Run every payment from the last 4+ years through JIL's 69-check Verdict Engine. Quantify your exposure to Fraud, Waste, Error, and Abuse. Recover what's recoverable. Regulator-ready output.
April 2026 - JIL Sovereign Technologies, Inc. - Patent Pending
Why this exists
Institutions don't know what they don't know. Most large financial, healthcare, government, and corporate organizations lose between 3% and 8% of annual payment volume to Fraud, Waste, Error, and Abuse - with no systematic way to detect, quantify, or recover it.
Internal audit teams sample. Regulators sample. Consultants sample. JIL audits everything. Every transaction. Every beneficiary. Every corridor. Every control point. At 1 basis point per historical transaction it is the first audit product that makes 100% coverage of 4+ years of payment history economically rational.
JIL's 69-check Verdict Engine (8 operational categories plus an emerging-threat intelligence category) is the only framework in production that treats Fraud, Waste, Error, and Abuse as a unified attestation layer rather than four disconnected compliance workstreams.
How it works
-
Secure data transfer
Customer exports 4+ years of payment records (wires, ACH, checks, cards, instant, crypto, CBDC) via encrypted bulk upload or authenticated API. Zero-knowledge pre-ingestion option via RISC Zero - no raw data leaves the institution's perimeter; only commitments and attestations are shared with JIL.
-
Parallel execution
All 69 checks run against every historical transaction in parallel. Standard throughput is 1M transactions per hour per shard. Multi-shard clusters scale to 50M+ transactions per hour. A typical 200M transaction audit completes in 48 to 96 hours.
-
FWEA Recovery Report
Every flagged transaction is categorized as Fraud, Waste, Error, or Abuse with quantified dollar exposure, root cause attribution, remediation recommendation, and a regulator-ready audit artifact (SHA-256 sealed verdict record plus optional ZKP proof). Output is Dilithium-signed and reproducible.
-
Remediation handoff
Findings export to existing case management and GRC tooling (Salesforce, ServiceNow, Archer, in-house platforms) with deterministic natural-language explanations from JIL's XAI layer for every finding - so recovery teams, internal audit, and regulators all see the same rationale.
The FWEA Framework
Every historical transaction is scored against four pillars. This is the core positioning of the product and the language we use with CFOs, CROs, Inspector Generals, and audit committees.
Deliberate deception for financial gain
Payments that should never have been made because the counterparty, instruction, or authorization was manufactured.
Example categories
- Pig butchering and romance-investment scams
- Business email compromise (BEC)
- Authorized push payment (APP) fraud
- Synthetic identity
- Deepfake voice / video authorization
- Magecart and card-not-present schemes
- Trade-based money laundering (TBML)
Relevant JIL checks
ID-001..007, PR-001..017, RG-001..011, SI-005..007, ET-001..002
Legitimate payments that shouldn't have been made
Duplicate, redundant, or inefficient disbursements that clear policy but burn cash.
Example categories
- Duplicate claims and duplicate invoices
- Overpayments and unnecessary top-ups
- Incorrect coding and upcoding
- Remittance mismatches
- Unused entitlements paid in full
- Redundant vendor payments across subsidiaries
Relevant JIL checks
HC-001..005, PR-002, SI-003
Unintentional mistakes in payment instruction
Operational mistakes where the institution intended to pay X but the system paid Y - recoverable, but only if detected.
Example categories
- IBAN checksum failures
- Wrong routing / account numbers
- BIC mismatches and invalid beneficiary
- Expired credentials accepted
- PO / invoice mismatches
- 835 remittance advice errors
Relevant JIL checks
SI-001..007, HC-004, RG-003
Payments that fall within policy technicalities but violate intent
Transactions that pass every individual control in isolation but form a clearly prohibited pattern in aggregate.
Example categories
- Structuring just below reporting thresholds
- Cross-jurisdiction smurfing
- Undisclosed related-party payments
- Sanctions evasion via hub routing
- Bust-out credit schemes
- Pig butchering grooming patterns
- CBDC restriction bypass
Relevant JIL checks
VL-004, VL-007, IT-004..006, RG-010..011, ET-001
Pricing - tiered monetization model
All pricing is per transaction in basis points (bps) of notional. The retroactive audit is priced at an 80% discount to real-time attestation because it runs in batch mode against historical data and does not consume the real-time SLA budget.
| Tier | Use Case | Price | Included |
|---|---|---|---|
| Starter Real-time |
Live pre-settlement attestation | 5 bps per tx | All 69 checks, XAI explanations, SHA-256 sealed verdicts, 99.95% SLA |
| Pro Real-time + Enrichment |
Live plus consortium intelligence network enrichment | 7 bps per tx | Starter + consortium signal enrichment (Enhancement K) + priority support |
| Enterprise Full Stack |
Live plus ZKP proofs, adversarial ML monitoring, XAI premium | 12 bps per tx | Pro + ZKP post-settlement artifacts + adversarial ML drift detection + dedicated compliance team access |
| Retroactive Audit (Discount) |
Historical batch validation of 4+ years of payments | 1 bps per historical tx (80% discount vs real-time) |
All 69 checks on every historical transaction + FWEA recovery report + regulator-ready audit artifact + 90-day findings remediation window |
| Retroactive + Real-time Bundle | Discount plus commitment to a real-time tier | 0.5 bps retroactive + real-time tier pricing |
Above + roadmap credit toward real-time implementation |
Example ROI - mid-size bank
A representative calculation for a bank processing 50M payment transactions per year over a 4-year audit window.
Ratios are illustrative. Actual findings vary by institution type, corridor mix, and control maturity. Healthcare payers typically see higher Waste exposure; government disbursement programs typically see higher Abuse exposure; wire-heavy commercial banks typically see higher Fraud exposure.
Who this is for
Get started
A typical engagement starts with a 30-minute scoping call, a signed MNDA, and a sample data structure review. First findings are usually in hand within 10 business days of data receipt.
Service: retroactive-payment-audit (batch mode of fraud-attestation-engine)
Related paper: Extended Fraud Intelligence - 69-Check Verdict Engine
Owner: Stanley Byrne (CIO)
Version: 1.0 - Published April 2026
JIL Sovereign Technologies, Inc. - Patent Pending - April 2026
This document is confidential and proprietary.