01Previously delivered Elucidate × Mastercard

Our existing engagement with Mastercard.

Elucidate is a BaFin-authorised, ESMA-registered financial crime risk infrastructure provider. This is not an introductory engagement: a Master Services Agreement is already in place with Mastercard, two Statements of Work have been executed, and our platform is integrated with the MMCP engineering team in their staging environment. The sections that follow detail the FI-assessment architecture deployed in MMCP staging and a sample EFI-scored report from that engagement — context for the subsequent product proposal.

// Commercial
MSA · in place

Master Services Agreement signed

Master Services Agreement executed with Mastercard procurement. No further commercial onboarding is required.

// Delivery
2 SOWs executed

Two Statements of Work delivered

Two scoped engagements have been completed with Mastercard, including the sample EFI-scored report shown in the following section.

// Technical
MMCP staging

Integrated with the MMCP engineering team

The Elucidate platform is integrated with the MMCP engineering team's staging environment. Connectivity and integration work is complete.

// Architecture deployed in MMCP staging

Entity resolution → enrichment → FI assessment → bank workflow.

The reference architecture currently deployed in MMCP staging. Open-source, proprietary, and Wolfsberg DDQ data feed a unified entity graph; third-party data sources enrich each entity; live transactional behaviour produces a risk score; the result is delivered to the receiving bank's workflow tool for green-list classification or manual review.

OPEN SOURCE DATA PROPRIETARY DATA WOLFSBERG DDQ REQUEST ENTITY RESOLUTION FI ENTITY ENRICHMENT FI graph + signal THIRD-PARTY DATA INTEGRATIONS BANKERS ALMANAC KYC REGISTRY D&B DUN & BRADSTREET OTHER EXTERNAL DATA API SYNTHESIS FI ASSESSMENT RISK SCORING TRANSACTIONS BEHAVIOURAL ACTIVITY CHECK API BANK SIDE BANK WORKFLOW TOOL GREEN LIST MANUAL REVIEW
// Sample deliverable

EFI-scored FI report — UBS, Mastercard test, high-score sample.

Output produced in the previous engagement: a single financial institution scored end-to-end against the Elucidate FinCrime Index methodology. The same scoring engine and data spine underpin the new product proposal that follows.

02  NEW PROPOSAL Proposed: the Payment Validation Agent. Addressing the RSP-routing problem raised in our most recent discussion.
02New product For Mastercard · Payment Validation Agent

The Payment Validation Agent, for MMCP's RSP routing.

Mastercard operates a network of approximately 200–250 RSPs, each maintaining its own risk appetite. Payments are currently bound to specific RSPs at the corridor level rather than routed dynamically, because no system component reads each RSP's policy at run-time. The Payment Validation Agent addresses precisely that gap: it ingests RSP policies, pre-validates each payment against them, and delivers the transaction with a routing recommendation already attached.

RSP-aware routing Real-time pre-validation EFI-backed scoring BaFin · ESMA Drafted justifications
// What we heard

Hard-tied routing across 200–250 RSPs.

Onboarding a new originating bank requires bilateral negotiation with every RSP. Origination from higher-risk markets, nested flows extending to the fourth party, crypto provenance, direct scheme connections into SEPA, and the recently introduced collections business with SME nesting all compound the operational complexity and manual workload.

Routing

Payments are bound to specific RSPs at the corridor level rather than routed dynamically, because no component reads each RSP's policy at run-time.

Onboarding

Every new originating bank requires bilateral negotiation with each RSP. Risk appetites are agreed pair-by-pair and do not propagate automatically.

High-risk origination

Markets including Congo and Rwanda, together with nested flows extending to the fourth party, demand case-by-case triage that does not scale operationally.

New complexity

Crypto provenance flows, direct scheme connections into SEPA, and the new collections business with SME nesting continue to accumulate within the same manual review queue.

It is so manual and heavy that something invariably goes wrong. — what we heard on the call
// Platform architecture

A schema-and-scoring stack, with the Payment Validation Agent on top.

A Common Data Layer feeds two scoring engines and a set of policy modules. A layer of autonomous agents operates on those signals under bounded autonomy. The Payment Validation Agent is the agent proposed for this engagement.

Layer 03 · Autonomous agents

Payment Validation Agent

Real-time pre-validation of every payment against the schema. The transaction arrives with a routing recommendation already attached, the evidence assembled across the underlying modules, and a drafted justification — in every case. When an RSP changes its appetite, routing recommendations update everywhere immediately, without each pair being renegotiated bilaterally.

Pre-validation Routing recommendation Drafted justification Bounded autonomy
Layer 02 · Scoring + policy

Transaction Scoring & EFI

A composite, explainable risk score per transaction and counterparty — combining static profile attributes (geography, sector, ownership, network) and dynamic behavioural data, in real time. The Elucidate FinCrime Index sits underneath as the institutional benchmark, with independent standing on both sides of any transaction.

Static + dynamic Explainable BaFin-authorised ESMA-registered benchmark
Layer 01 · Foundation

Common Data Layer

A single schema that feeds every upstream component — transaction data, counterparty data, RSP policy statements, and dynamic behavioural signals — normalised so the scoring engines and the agent layer operate on a consistent view.

One schema RSP policies ingested Real-time
// The agent in operation

Today's routing, then routing with the agent.

Step through the current hard-tied path one hop at a time, then switch slides to see how the Payment Validation Agent intervenes — pre-validating against RSP appetite within the network and enabling a routing path that the current corridor mapping would not select.

01 02 03 04 05 06 07 CORPORATE MSB SENDER BANK MASTERCARD RSP CORRIDOR-TIED RECEIVER BANK CORPORATE ORIGINATOR SIDE NETWORK RECEIVER SIDE
STEP 0 / 6
Slide 01 Today
RSP hard-tied to the corridor

The corridor's mapped RSP is the only available option. If its risk appetite does not match the payment, the flow is held or rejected — even where another RSP in the network would have accepted it.

Slide 02 Agent in-flight
RSP policies read at run-time

The Payment Validation Agent evaluates each RSP's risk-appetite statement against the transaction's composite score and pre-validates the payment prior to settlement. The routing recommendation is delivered with the transaction, not as a downstream triage task.

Outcome Routing
An RSP whose appetite absorbs the flow

The same payment routes through the same Mastercard rail to an RSP whose policy aligns with the transaction profile. When an RSP modifies its risk appetite, every routing recommendation across the network is updated in real time, eliminating the need for bilateral renegotiation.

// Proposed proof of concept

Validate on a single corridor — using your historical data.

Provide a 12-month transaction file from one of the affected higher-risk corridors. Elucidate will ingest the relevant RSP risk-appetite statements as written, process the file through the Payment Validation Agent against an indicative appetite profile, and deliver a comparison against your current decisioning.

OUTCOME 01

Where the agent agrees

Cases where the agent produces the same decision and routes to the same RSP, with a drafted justification attached — establishing baseline alignment between the schema and your current decisioning.

OUTCOME 02

Where it would have routed differently

Flows that an RSP with a compatible risk appetite could have accepted — payments not currently reachable through the existing corridor mapping.

OUTCOME 03

Where it would have released a held payment

Payments rejected or escalated by the existing manual review queue that the agent would have cleared, accompanied by the supporting evidence trail.

DATA · 12 months, one corridor SCOPE · Congo / Rwanda natural candidates RSP DOCS · appetite statements ingested as written DURATION · 1–2 months end to end from data receipt
// Next

Walkthrough of the Payment Validation Agent.

This document accompanies the walkthrough requested in our recent discussion. The session will cover the material above against your actual RSP appetite logic, followed by scoping of the corridor to be used for the proof of concept.

Schedule the walkthrough