Telecom fraud is a global problem, with almost 3% of the global telecom revenue lost to fraud in 2021 according to the CFCA. Our goal is to make communication fraud-free and to help operators and enterprises protect their customers.
The Fraud Management System (FMS) provides real-time Voice and SMS traffic protection – instantly detecting and countering fraudulent attacks. The FMS already protects 140+ clients and saves them $6M+ every month.
The FMS AI engine is trained on 174+ million live call attempts daily. It blocks up to 2 million fraud call attempts per day with new clients. After fraudsters realize that the route is protected they significantly reduce their efforts. The FMS analyzes traffic from many different service providers and thus can detect a large variety of fraud attacks rarely identified by individual operators. The FMS is signaling-based, which means that it analyzes active data in real time. It offers automatic detection of fraud attacks with immediate alerting and granular blocking.
The FMS is available as a cloud-based solution or it can be deployed on the client’s premises. The solution can be integrated with all network and signaling types. It works for MNOs, MVNOs, MVNEs, virtual PBX providers, fixed line operators, calling apps using SIP agents, transit carriers and enterprises.
The FMS AI-powered engine delivers an exceptional accuracy of fraud and spam detection – 99.92%. This metric is above the industry benchmark and was calculated using the standard formula:
(TP+TN) / (TP + TN + FP + FN),
Where: TP/TN – True Positives/Negatives, FP/FN – False Positives/Negatives.
The False Positive Rate of the FMS is below 0.00001.
Our in-house-developed ML algorithms adjust to the specific behavior of subscribers and partners from different regions.
The FMS detects the following fraud types fraud on inbound and outbound Voice traffic:
1. Inbound traffic:
- Wangiri Calls
2. Outbound traffic:
- PBX Hacking
- Wangiri Callbacks
- Enterprise Callbacks Fraud
- Conference Service Callbacks Fraud
- International Revenue Share Fraud
- Artificially Generated Traffic
The FMS is a customizable solution that offers clients full control over blocking policies via a user-friendly interface. Clients can monitor the fraud landscape in real-time on one dashboard. It’s easy to reverse active blockages in one click, store and edit blacklists, whitelists, and threshold-based rules. The System also has built-in reporting capabilities, so clients receive detailed reports according to custom settings on an hourly or daily basis. Once the FMS is set up, there is no need to review attacks on a case-by-case basis. It will automatically block attacks in real time before the damage is done.
We also offer a dedicated Security Operations Center (SOC). The SOC team offers full monitoring, technical and onboarding support and is available 24/7 for any client need – from adjusting settings to providing analytical expertise for a specific fraud case.
The FMS is constantly being upgraded to stay ahead of emerging fraud threats. In 2022, we launched SMS anti-fraud capability. It protects enterprises and operators from:
- Artificial SMS Generation
- SMS Phishing (SMishing)
- SMS Malware (SMS Hacking)
- SMS Originator Spoofing
- Prohibited Content
Another important update from 2022 is flash call detection capability. The FMS can accurately differentiate flash calls from other zero-duration calls. Operators can choose to block this verification channel or monetize it. Flash calls are detected instantly with machine learning algorithms and an up-to-date comprehensive database of flash call A-number ranges. The reporting tool supports potential monetization and business negotiations.
Additional Enterprise traffic protection capability was also added. If enterprises use non-automated voice services a built-in IVR module warns employees (sales reps/help desk assistants) when they dial high-risk numbers, protecting them from call-back schemes such as wangiri 2.0. If enterprises use automated voice services the ML algorithm accurately determines and sets granular blocking of B-number ranges related to fraudulent activity only. Thus, it prevents further calls to nearby B-ranges without affecting traffic to real subscribers.
Innovation and scientific approach lie at the core of our FMS. Half of our young employees responsible for building our FMS have Ph.D. degrees. Moreover, almost all of our FMS team members are involved in scientific research and working with real traffic.
In developing our FMS, we have implemented unique methods of fraud analysis based on time series analysis, traffic profile analysis, and machine learning majorly developed entirely in-house.
We have also developed an innovative system of case (fraud attacks) labeling. Machine learning algorithms then use the data labeled by our expert team for training. Our deep neural networks analyze 200+ complex parameters of fraudulent and spam calls. These parameters were developed in-house. This leads to an exceptional performance of fraud and spam detection methods.
The FMS undergoes constant quality control. We generate knowledge daily by analyzing all fraud cases detected in the past 24 hours. Each detected case is rated according to the following performance indicators:
- accuracy of fraud type classification
- accuracy of fraud-risk values set by the anti-fraud system
- accuracy of A / B range blockage
- time duration of fraud detection (from attack start till blockage), etc.
If any of the performance indicators are unsatisfactory, we immediately assemble a team and eliminate the cause of the problem through rigorous testing.
The FMS was developed by LANCK Telecom – a global player with 20+ years in the telecom industry, offering solutions for Voice and SMS traffic termination, roaming, fraud management, and communication solutions for enterprises.
The compatibility of the FMS with networks worldwide, the wide variety of fraud types it counters in real time, and outstanding detection accuracy of 99,92% and paving the way to a fraud-free future.