Dynamic AML Detection
Every payment screened as it happens, every alert with its evidence.
5detection rules, run on every transaction
1.2M+sanctions records from OpenSanctions and OFAC
Everyalert carries its evidence and a risk score
The problem
Money-laundering checks often run in overnight batches and flag far too much. Analysts drown in alerts and cannot see why one was raised.
This platform screens each transaction as it arrives, against live sanctions data, and every alert says exactly what triggered it.
The idea
1ScreenEach payment is checked against live sanctions lists as it arrives.
2DetectFive rules look for known laundering patterns.
3ExplainEvery alert carries its evidence and a risk score.
4ReviewAnalysts work the alerts on a dashboard, each with their role.
What was built
- Five rules. Sanctions screening, high-risk geography, structuring just under the 10,000 reporting threshold, unusual velocity, and money that goes out and comes back.
- Live sanctions data. Pipelines that load more than 1.2 million records from OpenSanctions and the US OFAC lists.
- Evidence on every alert. The matched entity, the corridor, the pattern or the threshold, with the reason it was flagged.
- Always available. A cloud database with automatic fallback to a local one if the cloud is unreachable.
- Access by role. Sign-in with Google, Microsoft, GitHub or Oracle; user, analyst, manager and admin roles; every sign-in logged.
- Test data on demand. A generator of realistic transactions with chosen risk profiles, so every rule can be exercised.
Built with