FraudStopper Frequently Asked Questions

Here at ECG, we're very excited about FraudStopper, our toll-fraud detection system. It is the only Fraud Detection tool optimized for VoIP service providers running BroadWorks, Metaswitch, or similar platforms.

You can get more details on FraudStopper on the FraudStopper page.


Here are some of the common questions we're getting.

What kind of fraud does FraudStopper stop?
FraudStopper is intended for the most prevalent form of Toll Fraud for VoIP Carriers. In most scenarios, someone has VoIP service with no SIP authentication, or very weak SIP credentials. For example, the password might be, literally, "password". Criminals discover this account via the Internet, then take over that user's account and place expensive outbound calls.

What are FraudStopper's thresholds?
Most of the simple Fraud prevention tools are based on simple assumptions about user's behavior. For example, many of them will alert if they ever see two concurrent International calls from one customer. Or they may alert if they detect more than 5 International calls in a day.

As long as these work, then that's great. But what about the customer who regularly makes a dozen International calls each day? You have to start building in exceptions.

Instead of building in a lot of exceptions, FraudStopper dynamically and automatically compiles a User Behavior Database of each user's behavior. So it knows that Fred makes 50 international calls every Tuesday, but almost none on Saturdays. And it knows that Granny never makes an international call.

What versions of FraudStopper are available?
FraudStopper comes in two versions. In both cases, FraudStopper is installed in your own network.

  • ECG Trained Analysts. FraudStopper with Human Analysts provides real-time, 24x7x365 monitoring for fraud on your system.
  • Your Own Analysts. FraudStopper sends alerts only to your staff.

How is FraudStopper priced?
FraudStopper pricing is based on the number of monitored users, and the type of analyst services you want.