inspkt Data Analysis
The system collects traffic data from predefined data sources, applies data profiling and analysis, detects and reports fraudulent cases on a timely basis, generates alerts, and executes actions onto the operators’ network elements to stop the identified fraud attacks and prevent future attacks from the same sources.
The detection mechanism works in real-time through our network monitoring probes, inspkt probes, or in near real-time if data is collected from the operators’ network elements.
The system features a machine learning algorithm which makes use of state-of-the-art machine learning models and techniques to ensure robust detection with very high specificity and sensitivity rates and cumulative learning capabilities. It’s a multilayered detection process that analyses callers’ behaviors in sequential stages, offering very precise and explainable results, making it the ideal solution for optimal decision making.
inspkt Data Analysis System is tailored for fixed and mobile operators and telecom regulators in order to detect:
- Wangiri attack
- Bypass fraud
- PBX hacking
- Call spoofing
- Robocalls
- Roaming fraud
- Subscription fraud
- SIM cloning
- Any type of IRSF fraud
- Operator’s internal fraud
- And other types of fraud
Key Features and Benefits
- Real-time fraud detection
- A dynamic rule-based engine to create an unlimited number of rules to handle any type of telecom fraud
- Automatic blocking of fraudulent numbers
- Automatic blacklist updates to prevent future attacks from the same sources
- Dynamic reporting engine with built-in and ad-hoc reports and live dashboards
- Seamless integration with third-party systems
- Adaptation to the fraudster’s “Modus Operandi” / behavioral change
- Detecting complex behaviors (human-like)
- Incremental learning
- Minimal to no human intervention for maintenance