Risk and fraud
In productionAlerts your compliance team can explain.
Sanctions screening on official lists, an AML queue with owners and deadlines, suspicious activity files with four eyes, and a risk center with 27 alert types on thresholds each operator sets.
What it does
Sanctions screening
OFAC, UN, EU and UK lists, downloaded every six hours from fixed sources and checked at sign-up, at KYC and before a withdrawal. A match freezes the account, and a stale list never counts as a check.
An AML queue with owners
Every open match has one owner in compliance and a 24-hour deadline. Dismissing a match needs a second person, and the database enforces it.
Suspicious activity files
Reports prepared with four eyes, never visible to the player, with a filing package for 11 financial intelligence units. Your team files them.
Risk center
Twenty-seven alert types, from deposit bursts, shared IPs and large withdrawals to stake jumps, bot-like round cadence and RTP drift, on thresholds each operator sets.
Geo rules
The country is resolved on the server, the operator's allow and block lists sit on a platform floor, and an unknown country is refused outside the demo.
- Coming soon
PEP and adverse media screening
Self-declared PEP status and enhanced due diligence work today. Screening against PEP and adverse-media data needs a data provider and is not connected yet.
By the numbers
Sanctions lists
5
OFAC SDN, OFAC Consolidated, UN, EU, UK
Alert types
27
Fraud score rules
10
Details
Lists you can audit
Each list is versioned with its SHA-256 and downloaded only from fixed hosts. When a list changes, only the new entries are screened again, in batches every five minutes. Matching combines Jaro-Winkler similarity, a phonetic key and the date of birth.
Monitoring for real money
Every 15 minutes, five typologies: structuring under a threshold, withdrawal after minimal play, rapid in and out, shared payment methods and third-party payers. It watches operators with real money only, opens cases and never moves money by itself.
Thresholds you own
Each operator sets its thresholds with a history of changes and can simulate a rule over the last seven days before saving it. Alerts are assigned, escalated, limited and resolved, and can be sent to Slack or email.
A fraud score from 0 to 100 is built from ten named rules and recomputed every night, so each point has a reason.
What is not there yet
Said plainly, so nobody plans around it.
- PEP and adverse-media screening, which needs a data provider.
- The multi-account identity graph: deployed and switched off.
- VPN, proxy and device-fingerprint detection.
- Filing directly with a financial intelligence unit: files are prepared, not sent.
Contact
Tell us what you are building.
Markets, brands, verticals. We show you the platform with your case in mind.