Fraud Prevention
Block Multi-Account Abuse
1:N face search, device fingerprinting, VPN/proxy/Tor detection, and synthetic identity detection to eliminate duplicate accounts and bonus abuse.
How it works
1
Identity Submission
User submits selfie and optional document. Extracted biometric and device signals are sent to the RoxVerify deduplication engine.
2
1:N Face Search
The submitted selfie is compared against your entire enrolled user base. Duplicate biometric matches are flagged in real time — no charge for deduplication queries.
3
Device & Network Fingerprinting
Device fingerprint is cross-checked against your user base. Shared devices, VPNs, proxies, and Tor exit nodes are flagged automatically.
4
Synthetic Identity Detection
Document data is cross-referenced for internal consistency. Generated names, impossible birth dates, and mismatched data are surfaced.
5
Decision & Action
Each check returns a deduplication score, list of matched accounts, and recommended action — block, manual review, or pass.
Key capabilities
👥
1:N Face Search
Biometric deduplication across your entire user base. Find duplicate accounts regardless of how many identity documents they use.
📱
Device Fingerprinting
Cross-device linkage identifies users who cycle through devices to create new accounts.
🌐
VPN/Proxy/Tor Detection
Network signal analysis flags users attempting to bypass geographic restrictions or hide their true location.
🤖
Synthetic Identity Detection
Catches fabricated identities that pass document checks by cross-referencing identity data fields for internal inconsistencies.
🏷️
Cross-Platform Flagging
Optionally share deduplication signals across your product family to catch users who abuse multiple properties.
⚡
Real-Time Decisions
Sub-2-second deduplication results with structured match data and recommended action — no manual queue required.
SDK & Integration
Drop-in SDKs for every major platform. Full REST API for custom integrations.
iOS SDKAndroid SDKReact NativeFlutterWeb SDK (JS)REST APIWebhooksNode.jsPythonJava
Ready to get started?
Talk to our team and be live in your test environment within 48 hours.