Easing Into Machine Learning for PSD2: An Interview with Our VP of French Sales
Latest posts by paulbookstaber (see all)
- Easing Into Machine Learning for PSD2: An Interview with Our VP of French Sales - September 13, 2017
- “How’s the View?” And Other Questions for Our New VP of German Sales - August 9, 2017
- New Revenue and New Risk, In a Payments World Powered by Partnerships - August 3, 2017
Jean-Damien Rubal has joined Feedzai to bring AI innovation to the French market. His arrival to Feedzai comes just as organizations are preparing for a post-PSD2 world, where new kinds of payments spell new risks – and new opportunities.
What do you predict in these months leading up to PSD2?
With 2018 fast approaching, PSD2 will oblige banks to open their infrastructures, and their customer data, to new fintech players. Consider Orange, the French telecom company that’s projected to gain 2 million new customers by launching a retail bank and allowing mobile payment transfers. Open APIs will allow for entirely new kinds of financial partnerships like this one. How can banks carve out and monetize their own spaces in this emerging and expanding landscape? That’s the kind of question echoing in bank offices across France.
What are the risks that PSD2 is introducing?
French banks fear new kinds of security breaches and data losses and corruption. More importantly, they fear losing control, because they’re losing the monopoly on customer information. Previously, banks had the benefit of operating with customers face to face. But as PSD2 brings a shift toward open banking, new innovators are arriving to the market that can do new things the banks can’t yet, like real-time money transfer. The resulting challenge for banks will be to build new business models that add value for their clients and cultivate their loyalty, while mitigating new risks at the same time.
How can machine learning help?
Machine learning will power this shift with all the intelligence we need to mitigate the risk associated with this change, because every time there’s a new way to pay, there’s a new way to commit fraud. The new focus is on identifying individuals without adding friction. Machine learning is the tool that can help banks make sense of the mountains of new data that will be generated under PSD2, so banks can weave customer authentication into the fabric of their new products and interactive services.
You mentioned adapting business models. What do you mean by that?
Many people are describing PSD2-readiness as going through a checklist to fortify their infrastructure. For example, our CEO recently proposed a checklist in the Paypers Open Banking report, describing the new decision sets that banks are grappling with as they examine customer liability and authentication.
But before ticking boxes off in a checklist, I advise banks to take a step back. This is the time to think about your business model, and consider new ways to position yourselves. Are there new customer segments to target? New value propositions to offer? Banks should also be thinking about new requirements to define and build.
It’s ironic that the times when it’s most important to take a step back are those times when there are so many details to look at up close. For example, banks are busy finding ways to provide a robust security environment, one that reduces customer liability, specifically for online payments. They’ll have to provide customer authentication that’s not painful, and they’ll have to perform due diligence on any customer complaints. This is all due by September, 2018. That’s a lot of boxes to tick, which means now is the time when it’s both most challenging, and most necessary, to think about reconsidering their business models.
Where should they start?
Banks will need to break down silos. Most people have multiple bank accounts, which means there’s a lot of different information about each individual that’s spread out in disparate places. The right platform can help banks correlate data so they can build complete customer views, even when the sources don’t normally “talk” to each other. Beyond that, the right banking aggregation platform can provide a dashboard to tell customers how they’re spending their money. This is the kind of value added service that a bank can now provide under PSD2.
So banks should begin by focusing on the right channels and striving to know their customers even better, so they can establish authorization and legal processes, control for AML and fraud, and make many other decisions. They’ll also need to put a system into place that can detect politically exposed people, and account for those who are blacklisted.
Why do you believe Feedzai is the right PSD2 partner?
What we’ve seen in France is a reflection of Europe at large. Initially, banks in France were protesting against PSD2. But now the large banks are recruiting data scientists to identify the best platform to create value out of their core asset: customer data.
The Feedzai difference is that we’re more than just omnidata and omnichannel. What we do is give you control. Our technology lets us train, test, and deploy models more quickly than our competitors, so we can solve your problem sooner than anyone. We also have the technology to process profiles of every entity, like card data, in real-time streams. Our competitors know what the card was up to in the past few minutes. But we know what it’s been up to in the past few seconds. That’s the difference between not detecting fraud and detecting fraud. Speed matters, and accuracy matters. We give you both.