Flag Fraud in High-Risk Corridors
“For each account, which sanctioned-country beneficiaries will receive wires over $25K?”
Book a demo and get a free trial of the full platform: research agent, fine-tune capabilities, and forward-deployed engineer support.
By submitting, you accept the Terms and Privacy Policy.
Loved by data scientists, ML engineers & CXOs at
A real-world example
For each account, which sanctioned-country beneficiaries will receive wires over $25K?
Sanctions screening catches exact name matches at transaction time. But it can’t predict which accounts are about to send large wires to OFAC-listed entities. Predicting these connections before the wire is initiated lets you pre-set blocks or require enhanced verification — stopping the violation before it happens. OFAC penalties reach $500K–$10M per incident.
How KumoRFM solves this
Graph-powered fraud intelligence
Kumo’s filtered link prediction restricts predictions to a specific subset: beneficiaries on OFAC sanctions lists receiving wires over $25K. It analyzes the account’s historical wire patterns, beneficiary geography, and correspondent banking relationships to predict that A002 will send a $32K wire to BN18 (Syria Import) — a violation waiting to happen.
From data to predictions
See the full pipeline in action
Connect your tables, write a PQL query, and get predictions with built-in explainability — all in minutes, not months.
Your data
The relational tables Kumo learns from
Accounts
| account_id | account_holder | risk_tier | relationship_years |
|---|---|---|---|
| A001 | Apex Corp | high | 5.2 |
| A002 | Trade Intl | high | 1.3 |
| A003 | Vega LLC | medium | 7.8 |
Wire Transfers
| wire_id | account_id | beneficiary_id | amount | timestamp |
|---|---|---|---|---|
| W001 | A001 | BN05 | 45,000 | 2025-01-05 |
| W002 | A002 | BN18 | 32,000 | 2025-01-12 |
| W003 | A003 | BN05 | 8,000 | 2025-01-10 |
Beneficiaries
| beneficiary_id | beneficiary_name | country | sanctions_list |
|---|---|---|---|
| BN05 | Iran Trade Co | Iran | OFAC |
| BN18 | Syria Import | Syria | OFAC |
Write your PQL query
Describe what to predict in 2-3 lines — Kumo handles the rest
PREDICT LIST_DISTINCT( WIRE_TRANSFERS.BENEFICIARY_ID WHERE BENEFICIARIES.SANCTIONS_LIST = "OFAC" AND WIRE_TRANSFERS.AMOUNT > 25000, 0, 30, days ) FOR EACH ACCOUNTS.ACCOUNT_ID
Prediction output
Every entity gets a score, updated continuously
| ACCOUNT_ID | CLASS | SCORE | TIMESTAMP |
|---|---|---|---|
| A001 | BN05 | 0.88 | 2025-02-01 |
| A002 | BN18 | 0.79 | 2025-02-01 |
| A002 | BN05 | 0.65 | 2025-02-01 |
Understand why
Every prediction includes feature attributions — no black boxes
Account A001
Predicted: 88% probability of wire to OFAC-listed BN05
Top contributing features
Wire amount to BN05
$45,000
38% attribution
Beneficiary sanctions_list
OFAC
28% attribution
Account risk_tier
high
17% attribution
Beneficiary country
Iran
11% attribution
Relationship years
5.2
6% attribution
Feature attributions are computed automatically for every prediction. No separate tooling required. Learn more about Kumo explainability
PQL Documentation
Learn the Predictive Query Language — SQL-like syntax for defining any prediction task in 2-3 lines.
Python SDK
Integrate Kumo predictions into your pipelines. Train, evaluate, and deploy models programmatically.
Explainability Docs
Understand feature attributions, model evaluation metrics, and how to build trust with stakeholders.
Bottom line: Predict sanctions violations before they occur. Pre-block or require enhanced verification on predicted high-risk wires. Avoid $500K–$10M OFAC penalty exposure per incident.
Related scenarios
Explore more fraud predictions
Topics covered
