Dear friends and partners!
We are so pleased to announce this year’s game-changing event. A few days ago, we launched the new 15th gen API JuicyScore. The release features a full set of new fraud markers. Our team enhanced parameters for remote access identification and social engineering. What’s more, we improved our technological stack to detect device-assisted virtualization and randomization. Besides, the new version comes with a set of improved IDX indexes.
What’s New in API 15?
- The response vector got 15 new variables.
- Scoring informativeness growth exceeded 30%, and indexes’ informativeness in the new API version increased by more than 55%.
- We managed to minimize 1st and 2nd kind errors for device authentication.
- 3 new technologies were added;;
- 18 technologies have made crucial improvements..
- 80+ technology abnormality and randomization tests were added.
- System performance grew by 25% thanks to a more stable architecture.
Updated and newly developed rules made it possible to ensure a tremendous growth of index informativeness along with 15 new API variables added. Below, you can watch the dynamics of index informativeness growth in the latest versions of our API.
Generic Score is a blend of models built upon IDX1-IDX10 aggregators along with a set of auxiliary variables.
IDX1: Stop Markers
IDX1 contains a combination of 40+ rare events. Each event represents a high probability of fraud through device technical manipulation or introducing meaningful abnormalities in the internet connection.
The variable comes with numerous means to detect device randomization. What’s more, it can identify techniques to interfere with their digital footprint. IDX1 detects the most harmful markers that define users’ inadequate fraudulent activities as well as internet connection markers.
Stop markers can be used together with other rules. Additionally, it can be used as a separate model’s component to prevent the risk of fraud and detect customer’s weakest segments. The risk level grows together with the underlying parameter. In simpler words, the higher the IDX rate, the higher the risk of fraud. High values can be used as filters to trigger automatic rejection.
IDX2: Users’ Behavioral Markers
An aggregated variable represents a blend of different users’ behavioral markers.
The JuicyScore vector contains dozens of markers related to the user’s behavior in one way or another. When designing this particular variable, the main challenge was to bring out stable markers united in a single aggregate. It helps to reveal high-risk segments that refer to high-risk fraudulent user behavior despite the online company’s location.
IDX3: Device Markers
IDX3 comes with a mixture of secondary markers of device-related abnormalities and risks.
Here, each abnormality itself can be a sign of potential risk while a combination of abnormalities, when triggered simultaneously, defines a specific risk zone. High index values can be used as filters to trigger automatic rejection.
IDX4: Internet Connection Markers
Here we have a blend of network parameters and abnormalities.
High index values can be used to detect and prevent the risk of fraud.
IDX6: Internet Infrastructure Quality Index
The Internet infrastructure quality index is highly recommended for evaluating credit risks as well as identifying the risk of social engineering fraud.
Besides, it can be used as the parameter of the credit scoring model. The index value greater than or equal to 2 reveals a more premium sub-segment in the application stream. It is generally associated with a better internet connection.
IDX7: Device Application Quality
The variable is available only via a mobile app. It provides an aggregate value of the apps installed on the applicant’s device.
IDX7 is recommended to identify credit risks and different fraud risk typologies.
IDX8: Aggregated Data Assessment from the Application
IDX8 is an indicator that reveals manipulations with application data.
The variable contains different parameters including multiple repeating phone digits or user’s locations as well as repeating device imprint.
IDX9: Applicant’s Mobile Device App Risk Level
The value of the variable shows the aggregated risk level of apps installed on the applicant’s device.
IDX10: Disposable Income Index
The index shows the population risk segment by the level of disposable income.
If the higher the variable value, the lower the fraud risk. We highly recommend using this parameter for evaluating operational and credit risks.
New API 15 Variables
Technologies skyrocketed to a new level triggering the growth of fraud scheme complexity. So, each year, we invest our financial and time resources in exploring, analyzing, and developing new variables and rules that would not only let us keep up but also outperform scammers.
Below, you can see examples of adding new variables and rules into API 15.
The above-mentioned indexes and variables represent only a small part of the bigger API 15 update. We realize the importance of constant enhancement of the markers’ set along with stop-factors that ensure momentum cost saving and loss reduction caused by fraud risks. Our team constantly works on technology improvements and innovations to provide our partners and clients with state-of-the-art instruments to evaluate suspicious and high-risk customers as well as prevent their fraudulent practices.