Payments embrace deep learning

As the market embraces technology to assist with regulatory compliance and better customer service, the sophistication of offerings available to end users has skyrocketed. PaymentEye spoke with René Hendrikse, Vice President & Managing Director, EMEA, Mitek, to find out more.

Onboarding verification checks seem like a painful necessity – but are also the first handshake between firm and customer. Do customers really lose interest in the firm if their onboarding experience is poor?

We’ve found that it’s more than losing interest, it’s losing the customer, and the potential revenue altogether.

Analysts and customers have found that up to 40 to 60 percent of consumers will abandon the on-boarding process when they’re applying for a new product or service. This can be a very costly result, given how much time and money goes into acquiring the customer, not to mention building out and promoting digital onboarding.

There are a couple of keys that we’ve seen that make up the most successful onboarding journeys: The financial institutions and payments providers who make customer experience and UI experts key stakeholders when building out their digital customer services are often the most successful. The identity verification portion of the onboarding process is vital to meeting the latest regulations, and it’s easy to see it as a burden and just another point for customers to drop off. But what we’ve found is that when companies can offer a better, more seamless experience, they get more customers through the onboarding journey – and those customers will also come away with a better impression of their service. So rather than seeing these checks as a negative, a good identity verification experience can be an opportunity to not only secure more customers, but also to start the relationship on the right foot.

Regulation seems to be putting payment processors and FIs in a rock and hard place scenario. Greater verification is required and yet consumers demand more seamless and consistent experiences. Which regulations would these be specifically?

Some of the most notable recent regulations are the Second Payment Services Directive, as well as the Fifth Anti-Money Laundering Directive.

Money laundering in Europe doesn’t only threaten the stability and reputation of the financial services industry, but it’s also a significant threat to the economy as a whole. And these types of crime are only becoming more and more complex, sophisticated and rampant. Action was needed on a supranational level. That’s what spurred the European Parliament and Council when they passed the 4th Anti-Money Laundering Directive (4AMLD) a few years ago [May 2015], along with the recent update in April of 2018 (5AMLD); it introduced a common framework for the proportionate prevention of money laundering or trafficking via financial institutions, payments and cryptocurrency providers. While the fight against money laundering is key to combatting financial crime, it also presents an issue for financial services who need to meet customer’s ever growing digital expectations.

Additionally, the recent passage of the Second Payments Services Directive (PSD2) is a big driver for increased verification and authentication of customers and end users. With PSD2’s requirement for Strong Customer Authentication, many transactions now are going to need to be verified based on at least two or more elements of authentication. These elements are usually categorized into three areas: What you know, what you have, and who you are. Meaning that to verify themselves, the customer will be required to prove who they are by way of knowledge (in a question and answer process), in the form of something only they would know; by showing the possession of something only they would have, (like an identity document or token of some sort); and lastly by inherence, something that the customer is, usually a biometric of some sort like a fingerprint or a successful face comparison.

Technology is obviously the differentiating factor in solving for meeting regulations and reducing costs. Am I oversimplifying the matter?

No – it really is that simple. We can all agree that consumers want a simple onboarding experience. It’s what they expect in today’s on-demand world. Plus, identity verification is mandated under European law. To solve for this, the digital twist to the most common and universal means for identity verification is to present an identity document for digital capture. We’re all accustomed to taking selfies and pictures on our phone. It’s become second nature. The images are then transmitted for analysis by complex machine learning authentication systems. This method is a much more cost-effective for an organization than to have to build and maintain an internal manual review team – and more convenient for consumers than asking them to visit a physical location to verify their identity. Both of the aforementioned methods of identity verification are not only expensive, they cause consumers to abandon the onboarding process in favor of a competitor who can offer a better experience

But is the selfie verification simply a gimmick in the age of social media?

It might sound like it, but selfie verification isn’t a gimmick in the least. While it might go by a trendy name more associated with social media, for verification purposes, it’s actually a simple way to enable facial recognition which is quickly becoming one of them most prominent security biometric tools for identity proofing. Facial recognition is used every day, around the world, at airports, or events and other public places to verify individuals and to keep people safe from terrorist attacks and other nefarious activities. Using it as a biometric comparison to accompany the photo on someone’s identity document is what proves the person presenting the document is truly the person they claim to be. So, taking a selfie isn’t necessarily a gimmick – it’s a universal mechanism of our contemporary global culture. It’s a behavior that most digital-first users are accustomed to.

Looking ahead, machine learning is tipped to become the emerging technology of choice for onboarding. What’s your opinion?  

There’s no doubt that machine learning, or more specifically “Deep learning,” also known as deep-structured learning or hierarchical learning, part of a broader family of machine learning methods based on learning data representations, which is different from task-specific algorithms is going to change the landscape in a lot of ways.

Because deep learning is so well suited to solving complicated image classification and object recognition challenges, it’s being used today in identity verification and facial comparison technologies. For example, in the financial services sector, banks are required to verify the identities of their customers in order to comply with anti-money laundering [AML] and know your customer [KYC] regulations. In the past, customers had to visit a bank branch and present their ID to an employee. Not only was that inconvenient for the customer but manual identity verification processes cost banks hundreds of millions of dollars to maintain, since they often had to make tens of thousands of customer calls every month to refresh KYC documents, or update incorrect or missing identity information.

But thanks to the rise of artificial intelligence, financial institutions can now use digital identity verification solutions powered by deep neural networks. New customers simply use their smartphone camera to scan their ID, snap a selfie, and submit them to the bank. Behind the scenes, deep learning algorithms analyze and instantly verify the authenticity of the ID and perform biometric facial comparison to make sure the person submitting the documentation is indeed the person pictured on the ID. Because of the computing power of these deep neural networks, the entire identity verification process is completed in just seconds and done entirely through the digital channel – saving the financial institution tremendous amounts of money, enabling them to comply with industry regulations, and deliver a much better customer experience.

This is just one example of the ways that businesses are using deep learning and artificial intelligence to automate processes, create efficiencies, reduce costs, and deliver better services to customers. As we see how beneficial these technologies are, the rate of adoption will only increase. And more importantly, so will the rate of innovation. The very researchers who gave rise to the deep learning revolution – individuals like Geoffrey Hinton and Yann LeCun – are already working on the next big advancements in machine learning.


Rene leads Mitek’s team in EMEA, where he excels at delivering Mitek’s world-class digital identity verification solutions to financial services, telecommunication, travel, gaming and other verticals to help them achieve their digital transformation goals. René has 20 years of experience leading business teams in the software and telecommunications industries. Prior to Mitek, René was Managing Director of EMEA at Lilee Systems and served as VP of Worldwide Sales and Managing Director of EMEA at iPass. He previously led Global Sales teams at IBM’s MaaS360 and Verizon Europe. René holds a BS in Information Technology, Telecommunications and Mobile Communications from Utrecht University, The Netherlands.

To find out more, check out the PaymentEye webinar on this topic, on Thursday September 6 at 2pm BST. Register here.

Related reading