Fraud Prevention

Risk Scoring

Based on the analysis of user activity, it is possible to identify typical fraudsters indicators. Strange account names, unusual authorization countries, use of virtual cards, frequent device changes. Each of these features separately means nothing, but together picture changes. All this can be analyzed comprehensively and assessed in terms of risk. We score not only users but also the IP-addresses, devices, affiliates, and even promo codes.

Useful against Account Takeover, Payment Fraud, and Promo Abuse risks.

Transactions Scoring

We can also develop transactional scoring. But unlike other vendors, in addition to the classic scoring rules, we use not only billing data but also their correlation with authorization and activity data. This approach allows us to identify more suspicious transactions with a higher confidence level. We work with custom billing systems and unusual payment pipelines.

Helps with Chargebacks Prevention, Card & Payment Fraud.

Bots Detection

Bots are not a problem when there are relatively few, but sometimes a large number of bots can be a black mark for investors and negatively affect reputation. We have collected an extensive bots database, identified their features, and trained the neural network. Using authorization data, we can now analyze new accounts and identify bots among them.

Also, fraudsters often use bots in Promo & Affiliates Abuse and Payments Fraud.

User Connections Analytics

Antifraud systems detect fraudulent accounts that differ from regular ones, but often this does not allow identifying all accounts affected by the attack. In this case, seeing the relationships between different users will be helpful. This approach is possible by analyzing geolocations, IP addresses, device IDs, activity time, used cards, etc. The more features a pair of users have, the more their connection is obvious. So, whether a massive hack or an affiliate attack, all involved users will be highlighted.

Useful for almost all fraud analytics, from Account Takeover to Refund Abuse.

Fraud Alerts & Response

Just detecting an attack is half the job. It is also crucial to stop it in time and build the proper response processes. Given our security background, we can develop correct alerts with minimum false positives, understandable playbooks, and automation elements. In addition, we can integrate fraud analytics with your systems up to automatic blocking of detected fraudulent accounts.