Purpose-Built to Fight Fraud

All the Functionality and Features That Risk Teams Need

Data Studio™

A powerful suite of tools that lets fraud and risk teams do the work of data scientists.

Machine Learning

Artificially intelligent algorithms that automatically adapt to your business patterns and get progressively smarter.

Big Data Scale

Capable of ingesting the largest of data streams from any source.

Whitebox Scoring

Clear human-readable explanations accompany scores to demystify the machine logic.

Segment of One

Individual behavior profiles at a hyper-granular level, no more loose-fitting segmentation.


Works on any channel, any device, no matter how your customers choose to engage.

Want to see it in action?

Schedule a Demo with Your Own Data

Our technology is interesting to read about, but it’s fascinating to see it work in real-time on your own business data.

Contact us today and we’ll schedule a demo using your own data. With the ground truth of actual chargebacks and other fraudulent behavior as the benchmark, a head-to-head comparison of Feedzai’s detection performance using past transaction logs can immediately demonstrate the value of our machine learning approach.

Contact Us

Data Studio

Risk professionals can now do the advanced work of data scientists using a guided, self-contained environment. Feedzai DS™ is the easy, straightforward way for risk teams to create advanced machine learning fraud models. DS is available for businesses large or small.

Whether you’re an analyst, business user or data scientist, Feedzai is built with all the applications you need to streamline and automate the machine learning lifecycle:

From extraction of features, feature engineering, model generation and evaluation, Feedzai’s intuitive application interfaces guides users through the rapid development of cutting edge algorithms.

Feedzai’s expert engineers, scientists and analysts use the applications daily to develop the cutting edge fraud and risk models that are embedded into the Feedzai technologies.

Data Science

\ˈdā-tə ˈsī-ən(t)s\

1 : an interdisciplinary field about processes and systems to extract knowledge or insights from large volumes of data in various forms, either structured or unstructured

2 : study of where information comes from, what it represents and how it can be turned into a valuable resource

3 : needed by organizations to manage the rapidly expanding dataverse

Machine Learning

Feedzai’s machine learning software recognizes what’s happening in the real world, how it’s different from what happened earlier, and adjust its responses accordingly. Machine Learning models can detect very subtle anomalies and very subtle signs of fraud without requiring new code or new configurations.

Machine learning is replacing rules-only systems which no longer work because today’s fraud patterns are evolving too fast for human-coded rules to keep up with.

Machine Learning

\mə-ˈshēn ˈlərn-tiŋ\

1 : a branch of artificial intelligence

2 : systems that automatically learn, predict, act and explain using data

3 : ideal for fraud management

Big Data Scale

Scale up as needed. Control and command the best of modern data technologies. Hadoop, Cassandra, NoSQL means horizontal scaling capabilities.

Handle the volume, variety, velocity and veracity of modern commerce:

  • Lightning fast response times: <20 milliseconds
  • Hundreds of millions of predictive metrics updated
  • Deep learning behavioral profiles maintained for years of history
  • Data adapters can handle any file format, structured or semi-structured
  • Horizontally scalable: add capacity as you need it
Only Feedzai scores all of your transactions in real-time0
Horizontal Scaling

\ˌhȯr-ə-ˈzän-təl ˈskāl-tiŋ\

1 : add capacity as you need it

2 : linear increases without reconfiguration or downtime

Segment of One

Each customer, device, IP, etc. represents a Segment of One™, and not a sample of many. Feedzai’s machine learning system maintains individual behavior profiles or signatures at a hyper-granular level, no more lose-fitting segmentation. This provides more accuracy in predicting fraud and risk patterns. If the data point exists, it can be profiled.

Once profiled, the machine models can distinguish normal behavior from anomalous behavior that is indicative of fraud. The ability to track in real-time billions of profiles means even the subtlest of patterns cannot go on unnoticed.

  • Track and profile anything: card, customer ID, device ID, IP address, email address, product SKU, etc.
  • Profiles are maintained in-memory, ready to be accessed in milliseconds

\ ˈseɡmənt \

1 : each of the parts into which something is or may be divided.

2 : a group sharing like characteristics or distinctive and identifiable behaviors

Whitebox Scoring

Clear, human-readable dashboards with text that explain the decision and scoring logic. Feedzai demystifies the machine logic and keeps customers, operators, regulators and management informed.

  • Semantic layer fuses machine logic with human intuition
  • Weighted risk factors and confidence levels
  • Live dashboards with composable elements


1 : clear evidence and clarity of the internal functions or mechanisms

2 : not murky or mysterious
Antonym: black hole, blank, blankness, emptiness, vacuity, void


More data, more mobile, more business models—more channels. Today’s consumers expect engagement that is always-on and everywhere. Modern businesses, in turn, need systems that are channel-agnostic, data knowledge that is liberated from silos, and technology that is accessible to all users.

Feedzai transcends silos and works anywhere:

  • Ecommerce, in-store
  • Mobile, desktop, tablet devices
  • ATM, in-branch
  • MOTO, petrol/AFD


1 : in all ways, places without limits

2 : provide the customer with a seamless experience