Credit risk and fraud risk, scored together — not in two separate systems.
The same Fintory rules engine that powers behavioral security for digital banks, reconfigured for SME lending: one score, continuously updated, from underwriting through the life of the credit line.
SME lenders run credit and fraud as separate workflows, on separate data, at separate moments. Each one is reasonable on its own. The gap between them is where the losses sit.
Bureau score and cashflow, checked mainly at underwriting.
KYC and ownership checks, done as a separate step.
Shell entities and split credit exposure across commonly-owned, co-incorporated companies slip through — because neither check looks at the other's signal.
Each layer contributes signals. The engine correlates them before any credit action executes.
KYC biometrics, ownership tree (UBO / beneficial owner), and incorporation timing of the applicant and every related entity.
Live open banking data — cross-checked against filed revenue where available.
Cross-correlates Layers 1 and 2 in real time into one Unified Risk Score, before any credit action executes.
This is the same 3-layer architecture used elsewhere on the platform for consumer digital banking — reconfigured with different signals for SME lending, not rebuilt.
Recently incorporated entities applying for credit — a weak track record is itself a risk signal.
Multiple related entities incorporated within a short window of each other — a pattern the ownership tree reveals but a bureau score alone would not flag.
Synchronized anomalies across linked accounts — an unusual transfer, a bounced payment and a shift in income sources happening together is a stronger signal than any one alone.
Scoring, Anti-Fraud, Compliance, and Fraud Intelligence each evaluate the case independently. The fusion engine correlates them in real-time — and when they disagree, it explains why.
6 weighted sub-components — Payment Conduct, Cashflow, Credit Report, Bureau, Ownership, Collateral.
Rule-based alerts (S1–S4 severity) from transaction monitoring and behavioral analytics.
Regulatory action rules — PSD2 SCA, AML high-risk jurisdictions, KYC thresholds.
Cross-institution consortium database. A confirmed match overrides everything else.
A case scores 89/100 (Clear) — but Fraud Intelligence flags a confirmed fraud link. The engine doesn't just pick one. It surfaces the conflict, explains the precedence, and blocks the facility.
The analyst sees the conflict, the override source, and the full decision trace — every time. No black boxes. Every block, verify, or approval is explainable to regulators.
Interactive What-If: Analysts can ask "What would change this from Block to Clear?" and get a simulated answer from the live data — before adjusting any rule.
Toggle what the engine observes on the left; adjust the rule thresholds on the right. Both feed the same score.
Enough signal to pause, not enough to act. Additional verification is requested from the customer; the line stays open while it is outstanding.
Configured visually in the no-code rules engine — risk teams adjust thresholds, not developers.
A bureau check tells you what was true on the day of underwriting. Open banking events, bureau alerts and ownership changes re-score the customer for as long as the credit line is active.
The same escalation ladder used in 360° Behavioral Security, mapped to credit actions.
A single flag, no ownership signal.
Route to the risk team. No action taken on the line.
Two flags within 14 days, or a recently incorporated entity.
Request additional verification from the customer.
Ownership risk, synchronized flags and a weak bureau score together.
Cap or freeze available credit automatically.
Every signal visible. Every decision traceable. Every rule adjustable — without writing code.
Cross-institution intelligence and real-time correlation catch what point-in-time bureau checks miss.
From application to fused decision in under three seconds — with a full audit trail behind every call.
Every score change, override, and rule edit is logged with actor, timestamp, and reason — ready for regulators.
Adjust thresholds, weights, and rules in a visual editor. No engineering ticket required to respond to a new threat.
Every module below is a working interface with real data. Open the case queue to see four scenarios, or dive into any module.
Deploy SME Protection & Risk Exposure on its own, or alongside the modules you already run.