
Nethone
Nethone is a Know Your Users (KYU) company that allows online merchants, digital lenders, and the payments industry to holistically understand their end-users and separate legitimate users from online fraudsters. By using cutting-edge machine learning technology, Nethone can detect and prevent card-not-present fraud, including protection against account takeover, bots, remote desktopping scams, etc.
Fraud Solution Profile
Specializes in cyber-security with a focus on preventing fraud in remote channels, e.g. payments with credit and debit cards or when it is related to identity or account takeover. While staying focused on the singularity of the client’s business, Nethone Guard can help lower clients’ rejection rates, chargebacks, costs of manual review, and risk of fraud, without touching the UX of their service–automatically and in real‑time.
Follow a sample user session timeline:
Step 1
A user logs in to your environment. The countdown begins.
Step 2
Nethone Guard gathers information about how the user keyed in the login credentials for the present analysis. What was the speed of the keystrokes? How does the left hand compare with the right hand? What about the mouse movement? We store the info for future reference and comparison.
Step 3
Your user’s session is compared with all of the user’s previous sessions through the lens of their attributes. It’s a combination of behavioral biometrics and the “cookie hijacking model.”
Step 4
After 1 second, Guard will have evaluated hundreds of attributes. After 2 seconds, it’ll be thousands.
Step 5
A graph of networks is built by matching transactions using shared attributes’ values. If it’s an account takeover, a picture develops. The probability that the user is a fraudster is calculated.
Step 6
Three seconds later, human readable and comprehensible recommendations appear in Nethone Panel. No inscrutable “black box” here–this is Explainable AI. The recommendations can even be used for reporting requirements.
How does Nethone Guard work?
Guard combines exhaustive user profiling which extracts 5500+ attributes for every user, a flexible API that is able to receive custom product/integration data from the merchants and integrates with partners who provide additional data enrichment.
Nethone Guard decision engine is a hybrid solution that enables ML models to be combined with more static logic. Nethone is the author of its whole modelling stack allowing for constant improvement of algorithms and customization of the decisioning setup to suit the clients’ needs. An important part of Nethone’s value is the direct relationship of the Data Scientist with the client. A dedicated Data Scientist takes care of tying merchant’s KPIs directly to the decision engine logic.
Nethone technology is known for its:
- Easy integration
Even large eComm sites can be up and running in less than 2 weeks.
- Iterative upgrades
Access the latest technologies, plugin 3rd party APIs at will or send historical datasets to power Machine Learning.
- Explainable AI
Receive human readable recommendations complete with the reasoning behind the decision. It’s ELI5 methodology, not the typical AI black box. Many clients need to understand why a decision was made for reporting requirements. We can pinpoint that a particular model made a particular decision.
Farfetch
Azul
VTEX
Grover
Booksy
Wonga
Ecommerce, Financial Services
Fraud Platform
Account Takeover, Loyalty or Promo Abuse, New Account Fraud, Payment Fraud
Behavioral Biometrics, Machine Learning, Rules Engine