Easy Integration

Designed to Empower Developers

Simplified access to sophisticated algorithms.

Engineering an advanced machine learning platform that can deliver powerful performance with reliable results requires deep expertise in technology and data science.  Feedzai’s out-of-the box platform delivers these capabilities through high capacity sockets or easily accessible REST API.

With access to potentially thousands of external data sources, Feedzai can create 360-degree view of customer at hyper granular level to detect fraud with greater accuracy.  Transactions can be enriched using:

  • Device Fingerprinting
  • Entity Linking
  • Pattern Detection
  • Event Velocity
  • IP Tracking

Data is secured behind firewalls in PCI DSS compliant data vault, and can only be accessed via SSL.

Power up Machine Learning in 4 Easy Steps



1

Seed & train

Send data to the system

Each business is unique. Fraud models can be trained using data from historic transactions, black/white lists and other data that is specific to your business. This could also include non-transactional data like SKU, employee lists, POS data etc.

2

Track user activity

Track User device and browser information

In addition to receiving transactional data for fraud scoring, Feedzai also helps merchants track behavior before and after login to detect and prevent identity fraud. Integrating with Feedzai’s device fingerprinting technology is simple and can be invoked by embedding code snippet in all the webpages where customer interactions are expected.

3

Score

Score transaction in real time

Feedzai can help you risk score at any point in the transaction journey, either in the pre- or post-payment authorization flow. Once the customer completes the order, the transaction is sent to Feedzai through the main scoring API call.  Depending on the score and recommendation provided by Feedzai, you can choose to accept or decline the transaction.

4

Send feedback

Send your feedback to Feedzai

Feedback cycle is a critical step in a fraud prevention process.  The more feedback the models receive, the more accurate they become and the more fraud they detect.  When analysts make a decision using the alert manager,  feedback about that particular order is automatically sent to Feedzai. However, when a future event on an order changes its state, there will be no alarm to trigger in real-time.   This information can be still be transmitted to Feedzai manually in the alert manager interface or can be automated through an API.

API Documentation

Feedzai offers easy integration of payment and user behavior data through simple REST API. You can send payment information along with custom user actions (e.g. login, logout, web tracking) to the scoring engine and receive a risk score. For further details refer to the API documentation.

View Documentation

Get Started

Choose the API to integrate into any platform or plug-ins.