Intelligence - Overview
Telesign Intelligence helps you gain a comprehensive understanding of the phone number's risk profile by obtaining a risk score for the current transaction. You submit a phone number, and optionally an IP address and/or email address for the same user, to the service. The individual identifiers are each evaluated, and then a score is returned telling you how risky the current transaction is.
Along with the score, you can also get a set of insights reason codes that provide more granular and contextual intelligence on the digital behavior driving the risk score.
How it works
Evaluates static attributes
The service evaluates phone number, IP, and email data attributes including phone type, telecom carrier, account ID, device ID, VPN/proxy use, and domain related to the current transaction to identify potential fraud risk.
Evaluates behavior patterns
It also evaluates anomalous traffic behavior patterns and usage velocity, as these may raise red flags.
Examples of suspect behavior patterns include:
- Passcode requests are received in five different languages from the same number in the same week, from IPs in multiple countries
- If a range of numbers are seen repeatedly on one or more web services, it may be a sign that a digital profile is being shared.
When we identify suspect behavior patterns like these, the risk score increases accordingly.
Adapts to your needs
The Intelligence machine learning model can also evaluate custom labels and traffic provided by you to adapt to your scoring requirements. By combining all these intelligence sources, we can help you prevent attacks such as International Revenue Share Fraud (IRSF) and other types of scams. If you choose to use these custom features, you can tailor the service to look for scams that specifically affect your business and traffic.
NOTE:
To add customization to your instance of Intelligence, contact a Telesign expert.
Documentation by category
Reference
- Submit a phone number to evaluate - Obtain a score for this phone number, as well as other relevant information.
Docs
Updated 2 months ago