Ever wonder how a financial transaction is evaluated for risk when a machine learning platform is used? Wonder no more; we’ll show you!
Watch the video below to gain insight on how machine learning scores financial transactions for risk.
Video Transcript
It takes 300 milliseconds for the human eye to blink. It takes 13 milliseconds for the brain to process visual information. In less than 3 milliseconds, Feedzai can evaluate thousands of decisions to score a transaction in real-time. How?
Meet Diana. Diana is a data scientist who trains and evaluates machine learning models using Feedzai’s platform. The models she builds score the risk of a transaction and automatically identifies financial crime—all in real-time.
When a transaction enters the system, it is evaluated using sophisticated machine learning models to identify patterns that are not obvious to the human eye.
First, Feedzai extends the original transaction information with known customer behaviors and profiles. Next, external data points are added and Feedzai’s data consortium further enriches the information.
We then use machine learning models to assess the risk of the transaction and detect financial crime patterns. Additional rules and lists can be used to fine-tune the decision to approve, reject or review.
The result is a decision that has clear, white-box explanations that provide documentation and transparency
That’s when Rich, the risk analyst comes in. Once the transaction is scored he’s able to see the fraud alerts in a convenient list and can click inside the alerts to learn more about them. If something looks suspicious he flags the alert for Sonia, the senior data analyst.
Sonia uses Genome, Feedzai’s AI-powered visual insight tool. With Genome, Sonia can easily investigate patterns and make data-driven recommendations to uncover additional financial crime.
Feedzai’s artificial intelligence incorporates Sonia’s decisions back into the model in a closed feedback loop that continuously improves performance. With Feedzai’s Automated Machine Learning capabilities, teams can deploy new models – across any channel and geography – to manage risk with agility and speed. And that makes executives happy.
Feedzai. One platform to solve financial crime.
Share this article:
Jaime Ferreira
Jaime Ferreira is the Vice President of Risk & AI at Feedzai. His team collaborates closely with customers across various sectors to deliver robust fraud and anti-money laundering detection. Jaime's ambition is to create a safer world with less fraud, scams and money laundering.
Related Posts
0 Comments22 Minutes
Scam Liability Shifts: Global Banks Reach an Inflection Point
Who should pay for scam losses? How much should they pay? What types of scams are…
0 Comments10 Minutes
Could a Scams Lawsuit Shift the Liability Tide in the US?
The United States has lagged behind other countries regarding scam liability…
0 Comments13 Minutes
A Fraud Model to Reduce Customer Friction and Build Trust
Trust between banks and customers is a two-way street. First, banks must trust that the…