Experts from Featurespace, TSYS a Global Payments Company, Mastercard, and Worldpay from FIS share their solution requirements

Scaling for value added services 

In part two of this blog series, our experts shared the commercials and value propositions of their solution offerings. 

  • Dave Excell, Founder at Featurespace  
  • Dondi Black, SVP, Chief Product Officer at TSYS 
  • Liam Cooney, Vice President at Mastercard 
  • Mina Khattak, Director, Crypto & Emerging Business, Worldpay from FIS 

In part three, these payments powerhouses detail the solution requirements of the technology underpinning their fraud and Anti Money Laundering (AML) product offerings. 

Excell: “It’s one thing to have the technology, but there’s also the challenges of thinking about how do we deploy at scale and in real time in these mission critical platforms that we all run on behalf of businesses and customers.” 

Data, technology, and experience 

Khattak: “When deploying FraudSight at Worldpay there were three key considerations. The first is data, the second is technology and the third is the experience.  

From a data perspective, the robustness of your fraud score and your fraud model is a function of how much data throughput is coming through the system and what’s the quality of that data. With FraudSight we see 40 billion transactions annually across both card present and Card Not Present (CNP). We see customers sometimes just wanting to purchase our fraud score because it’s reliable and it’s accurate. So having high quality data is the first thing you have to solve for and that’s why it is advantageous as an acquirer to offer a fraud tool to minimize fraud, because you do have very high-quality data.  

The second piece is technology, both building a tool to analyse the data as well as from an integration perspective. We use a proprietary machine learning model that uses Adaptive Behavioural Analytics to be able to create robust rules around creating a tiered system of high, medium, and low risk fraud. The extent to which you can create a API driven, no integration solution will definitely enhance the rate at which consumers will adopt that, because integration creates a lot of friction.  

For experience, if fraud is detected and the merchant needs support or advice on strategies to mitigate that fraud in the future, or detect fraud in a more robust way, you need data scientists. You need IT consultants, and you need fraud analysts that have experience in the acquiring and issuing space to be able to devise those strategies in collaboration with merchants. You really need that engine with both data and technology, and a team that has experience in order to have maximum impact in this space.” 

Global delivery of fraud prevention  

Black: “We cover a number of segments, we’ve got the financial institutions fintechs, corporates, businesses, and commercial customers. In many cases we service them globally. The ability to create orchestration across their portfolios to share that data and leverage that in real time, and the ability for that model to learn and retrain itself in real time is critically critical.  

The orchestration comes in with sharing those learnings across those portfolios, which is key because very often you find some of the same anomalies. Today in a lot of cases, a lot of fraud systems and tools are really disaggregated across different portfolios, so the ability to pull that together and orchestrate that so that you are leveraging the learnings that you’re using across multiple portfolios is, incredibly, incredibly important.  

The journey that we are on to move our services to the cloud, and the increased value in terms of resiliency that comes with being able to offer Foresight in a cloud environment is critically important to our clients as they think about the real-life impact of resiliency needs and disaster recovery in ways that maybe we didn’t think about a decade ago.”  

Flexibility in fraud and AML products 

Cooney: “It’s a challenge and an opportunity for us when we think about scale and Mastercard. We work with more than 20,000 financial institutions and 30 million merchants across 210 countries and territories. So when we’re developing a fraud or AML solution, we need to make sure that that solution works, not just for one customer or in one market, but it that it can deliver value globally and do what it’s intended to do across that entire flow.  

Obviously, that’s challenging and there’s lots of complexity involved in it. But over the last several years we’ve gotten pretty proficient at working around some of those constraints and deploying these solutions actually pretty quickly at scale. A solution that we developed that that helps provide visibility into Fiat:crypto and crypto:Fiat transactions we adapted and built that product and deployed it within a year, which is pretty impressive. 

Both solutions and products that we design and develop need to be flexible and so does the underlying technology that supports it. The reason being that fraudsters and criminals are flexible. If we’re effective in shutting down certain tactics and certain attack vectors, they’re not going to give up and collect unemployment. They’re like water. They’re going to look to find the path of least resistance. The solutions that we deploy need to be able to adapt to changes in behaviour, pick them up quickly, and then put in solutions that will continue to deliver value to the ecosystem.  

The whole Money20/20 conference is about how payments and underlying technology evolves. So does fraud, and so does AML. Underlying solutions need to be able to be flexible enough in their design to be able to adapt as payment behaviours and payment types evolve overtime. And I don’t think that that’s ever going to change.” 


Part 1: The Next Big Thing in Payments as a Service? Fraud and AML

Part 2: Product Design for Fraud Prevention and AML in Payments as a Service

Part 4: Current and Future Fraud and Money Laundering Trends

Part 5: Machine Learning in Fraud and AML as a Service