Welcome to the Intelligence Hub
No two users are the same. That's why our solutions are built individual by individual.
Is the device making the transaction, the one that is associated to that user?
Based on our recognition and trust model, we collect key aspects from across the web browser and/or device making the transaction. We then combine this with malware and bot analysis to ensure that the said device isn’t compromised or being remotely controlled.
Mobile Device Fingerprinting
We collect data related to the features of the mobile device (e.g. hardware, operating system, installed apps) to indicate device identity.
Mobile Device Anomaly Detection
We check for signs of malicious activity or risk to the mobile app session e.g. root, jailbroken, tampering, hooking, emulation.
Application Malware Profiling
We scan the device and hashes of installed apps to check for known vulnerabilities against multiple malware engines.
We collect data related to the features of the browser (e.g. hardware or browser information) to indicate device identity.
We check for malware instances through data collected about the web page application.
Whitepaper: Bot Wars
How future AI powered BOTs will work against each other to mimic human behavior
In this whitepaper, we outline the threat posed by malicious bots, how they have matured and evolved with time, and what steps can be taken to mitigate the damage they are capable of inflicting.Download
Does the location at the point of transaction / time of day match the user’s typical behavior?
Today, mobile technology is ubiquitous and thanks to this widespread use and availability of GPS, WiFi, Bluetooth, and other device-reading sensor technologies, geographical location can play an important part in the identification process as another opportunity to passively confirm identity, helping to eliminate friction for the end user.
Providing that the user has consented and, staying true to our privacy principles - using obfuscated data, we confirm if the user’s location at the point of request aligns with their typical behavior. If it does, the they can get on. If it doesn’t, then the request can be refused, or additional authentication can be requested.
Whitepaper: Location-Based Authentication
Asserting user identity using geo-location and machine learning.
In this whitepaper, we discuss possible approaches to location-assisted authentication and introduce the machine learning-based solution for behavior-driven authentication that we have developed in-house at Callsign.Download
Is the user is who they say they are?
It’s no secret that the costs of fraud keep on mounting. Many current authentication methods whilst initially seeming secure, degrade over time, presenting the bad guys with new attack vectors. Behavioral analytics can play a major role in fraud detection, using passive analytics to help verify a user’s identity without adding friction to the journey.
We use state of the art machine learning techniques, to verify that device usage is unique to an individual, building a unique understanding of user behavior overtime. Taking this approach, we can prove identity with a greater degree of accuracy, rather than relying solely on static authentication that traditional biometrics and knowledge-based authentication brings.
Whitepaper: Intelligent Swipe Authentication
Using swipe behavior to provide a secure and frictionless method of authentication.
In this whitepaper, we outline the architecture of our swipe model, which makes use of AI and a range of sophisticated mobile behavioral data to verify the authenticity of a user.Download
Identification based on recognition & trust
We look at our data in modalities of device, location & behavior, before combining them using flexible weightings based upon its significance, such as when “unique” behavior is recognized. We then look at how trustworthy this information is, conducting ongoing passive analysis across various data points, learning the detailed nuances of each user profile.
By considering multiple data points during our predictions and analyzing those in combination, we can build up a unique identity profile for each of your users, spotting relationships between data points that traditional vendors would not.