Transaction Fraud

Stop Chargebacks Before They Happen

As digital banking, fintech, and social commerce businesses flourish, so do fraudsters and their sophisticated attacks. Gone are the days where a single attacker uses a single stolen credit card to make a quick score. Financial fraud has become a professional enterprise, with a complete ecosystem of fraud-as-a-service stealing over $16B per year. Using an unsupervised machine learning approach, DataVisor’s solution can uncover the most complex fraudulent activities before any monetary damage is done.

Request Trial

Common Attack Techniques

IP obfuscation

Fraudsters utilize proxies, VPN, or cloud-hosting services to hide their tracks from IP blacklists and rules-based systems

Disposable email addresses

Attackers use disposable emails to mass register accounts to carry out credit card testing

Device obfuscation

Fraudsters utilize mobile device flashing, virtual machines and scripts to appear as though the login events are coming from different devices

Account incubation

Fraudsters create and incubate fake accounts for a period of time to evade rules on account age

DataVisor Online Fraud Report

Learn about the techniques fraudsters use to evade detection

Download the report

Why Unsupervised Machine Learning for Detecting Fraudulent Transactions?

Where there’s money, there’s a fraudster trying to game the system. Despite advances in fraud detection techniques using supervised machine learning, fraudsters have learned how to appear legitimate to these trained models by continuously adapting their techniques. Powered by the latest big data technologies, DataVisor’s unsupervised machine learning algorithm takes a global view of all the transactions and accounts together to uncover highly correlated, suspicious clusters of fraudulent activities. When viewed in the full context of all accounts, a fraudulent transaction that once appeared legitimate now sticks out like a sore thumb. Even better, DataVisor’s unsupervised approach doesn’t require delayed chargeback labels from being able to find new attack techniques, thereby shortening the window of exposure to these attacks.

Unknown threat protection

Detect new categories of attack campaigns without any training data or labels

Unparalleled accuracy and coverage

Uncover linkage among malicious accounts to catch all members of the crime ring

Early detection

Detect malicious accounts at account registration to prevent chargeback and financial losses

Key Features

Real-time detection

Integrate using our Results API to receive detection results with sub-second latency to prevent fraudulent transactions before they happen

Event timeline visualization

Discover sequence of incubating activities such as registration, logins and account transfers to quickly discover patterns of suspicious behavior

Account linkage visualization

Show linkage and shared attributes among malicious accounts to visualize the entire crime ring

Event timeline view to visualize pattern and sequence of suspicious behaviors

Bringing Real Value to Our Customers

CASE STUDY

DataVisor Fights Fraud with Fortune 500 Financial Institution

This Fortune 500 financial institution suffered huge losses in fraudulent credit card transactions. Read this case study to learn how DataVisor help saved the financial institution millions of dollars per year.

Read Case Study

Getting Started

Want to get started and find out how DataVisor can help find malicious accounts hiding inside your online service? Request a security assessment today!

Request Trial