Last month, according to a Forbes article, you have recently broken the amazing threshold of processing USD 1 billion in payments volume per day. What are you doing right?
Yes, this is definitely an important milestone for us. In fact, with the expansion in our business we expect to be processing USD 2 billion in transactions daily in a few months. What we are seeing right now is that our technology is now running as a fundamental risk engine across a huge scale of the business of our key customers. As a consequence of our growth were now one of the 50 fastest growing startups in Europe, and the only one in Portugal.
Our customers are deriving strong business value and expanding their use of our technology across their business. One of our customers, an innovative American Retail Bank, has been able use Feedzai to increase approval rates on new accounts by 70% and reduce the average time to approve applications from days to minutes. Another customer, a global payments company, is continuously expanding their use of our platform across risk scoring of debit card payments to merchant protection services and more.
What is driving this is not only our superior accuracy but also our speed at scale; our average transaction processing time is just 2 milliseconds and we are running tens of millions of transactions daily. The beauty of this is that, with machine learning, the more data we process the better we get, ensuring that our clients provide exceptional customer service at ever lower levels of risk.
Why is machine learning so important for fraud and risk?
The great opportunity today is that as a society, we are generating ever greater amount of data such as social, location, mobile and not just consumers but merchants and businesses as well, and enabled also by that fact that more and more of commerce continues to become digital enabled. We can now use machine learning technologies to enable money lending to a customer that lacks a long credit history by quickly assessing the true risk, it means that commerce can truly work at Internet pace instead of being stuck in manual human verification processes.
At the same time fraud exposure is going up because legacy systems are not able to keep up with the increase in computational power available to fraudsters. Current fraud models are at many times rules based, and other times only updated after a lag of 6 to 12 months. A real-time machine learning system can quickly adapt to changing fraud patterns and learn from both historical and multitude of real-time data to combat fraud and reduce the costs of continually keeping up with new vectors of attack.
What is your unique value prop in the risk and fraud prevention space?
Great question, our roots as a company are really based in real-time analytics at big data scale, what this means is that not only do we deploy state-of-the-art machine learning technologies, we are able to do that for large volumes of data at high velocity.
So what does it mean for our customers?
1) Great performance that improves; in the same time legacy solutions even go-live, we are able to learn and improve performance. Offering much greater accuracy and precision.
2) We are quick to deliver value; legacy providers take a long time to implement and are only able to support fraud and risk models updates every 6-12 months. We can get up and running fairly quickly and our specialized data science environment enables new model updates in weeks.
3) We are truly omnichannel; we are able to be data agnostic, and work across the entire commerce ecosystem. This means that the same way commerce is omnichannel, so are we. Not just that, but when you add in data from different channels, it only makes our models better.
4) Our results are actionable; Some machine learning based solutions are known to suffer from a ‘black-box’ problem, where you get an answer but don’t really know why. We are able to provide ‘white-box’ explanations as to why we produced a certain answer. At the end of the day, there is still going to be a person that gives the go-ahead and with this she has much greater confidence on making the right decision.
Seems like business is good, what are you doing next?
Yes it is. We are making thoughtful investments across sales, marketing, product management and delivery teams. We are doing this in the US where we have offices in Silicon Valley and New York City and expanding further in Europe, where we were founded and is a market we know well too. Europe is hot again and is “what’s next” for many companies looking to expand globally, the rise of the Lisbon tech scene is really evident of this. For example, the Web Summit conference has now moved to Lisbon and will attract 50,000 attendees. Everyone is looking at Europe as the next place to be, especially in the FinTech space.
Who are you competing with?
This is an interesting question, we see ourselves as part of the disruptive change that artificial intelligence is having everywhere. What Amazon, Netflix and Facebook do with AI for their respective businesses, we use to fight fraud and manage risk across the entire payments ecosystem. However, fighting fraud is unique because we don’t face ordinary competitors, instead we face evolving adversaries that are professional criminal organizations who purposely don’t play by the rules. This changes the stakes considerably.
At the same time, in many cases our customers actually use us in conjunction with their existing systems, where were really additive. Here customers see us as a better way to upgrade to machine learning and big data driven solutions and “fight fire with fire” in the battle against fraud.
About Nuno Sebastiao
Nuno Sebastiao brings his experience in infrastructure and services as well as corporate management skills to Feedzai. Previously, Nuno led the development of the European Space Agency Satellite Simulation Infrastructure. Prior to his tenure at the European Space Agency, Nuno was a co-founder at Evolve Space Solutions, a services company working in the Aerospace domain. Nuno holds an MBA from London Business School.
About Feedzai
Feedzai is a data science company that uses real-time, machine-based learning to help payment networks, banks and retailers prevent fraud in omnichannel commerce. Feedzai’s fraud science technology fuses machine learning with human intelligence to power payments systems globally for customers in North and South America, Europe, and Africa.
Every day we send out a free e-mail with the most important headlines of the last 24 hours.
Subscribe now