Q&A – The Next Generation Data Platform for Financial Crime

An Interview with Dr Richard Harmon, Managing Director, Financial Services, Cloudera

Financial Crime is one of the biggest costs for financial institutions with recent research by Refinitiv indicating that global financial services (FS) organisations lost more than $1.4 trillion in 2018. Another Refinitiv report indicates that 97% of FS institutions believe that technology can significantly help with financial crime prevention. 

This interview with Dr Richard Harmon, Managing Director of Financial Services at Cloudera, explores some of the key challenges and approaches to tackling financial crime. Dr Harmon highlights the way some banks are innovating their approach with technology, data and analytics in order to combat threats in areas like Anti-money Laundering (AML), Fraud, Insider Surveillance, and Cybercrime.

Dr. Harmon, can you tell me a bit about your background and what you do here at Cloudera?

I’m the joint managing director of financial services at Cloudera working alongside Steve Totman. I’m based in London and Steve is based out of New York. In terms of my background, I spent more than 25 years in Capital Markets, particularly in Risk and Quantitative Analytics, working with firms such as Citi, Bankers Trust, JP Morgan, Bank of America and Blackrock.

I’ve been with Cloudera a little over three and a half years. I spend a lot of time travelling, visiting financial services customers all around the world, evaluating what they’re doing with data and where they’re finding successes leveraging our capabilities, sharing experiences and successful use cases executed by other firms, and understanding what their challenges are.

As industry heads, we support our sales teams. We help identify the future trends that we should focus on in terms of use cases and partners. As an innovative platform provider anything we do on a specific use case is done directly by the customer or through an application owned by the partner. So, from that side, our partner ecosystem is critically important.

Can you tell me more about Cloudera’s credentials when it comes to working with global financial institutions?

We have an envious footprint, with over 525 customers in financial services, including 82 of the top 100 banks. So with that kind of a footprint we cover a significant part of the landscape in enterprise financial services. That gives us a really strong view of what our customers are doing, what their challenges are and how different regions have different priorities. For example, the Asia Pacific (APAC) region is incredibly mobile-driven, everything is being digitised, whereas Europe and the US are behind this curve. 

The other aspect, which I think is a real differentiator for us, is the fact that we are an open source platform. The innovations are not just coming from Cloudera, they are coming from a wider global data and analytics community, which is especially true in the area of machine learning and artificial intelligence (AI). In that space, innovation is being driven by major universities, by academics, by major institutions of all types and a lot of this innovation is put into the open source space. So the fact that our technology enables people to leverage all of that innovation globally gives us an incredible position in the marketplace and an ability to support our customers for all of the next generation activities they want to build.

You mentioned to me recently, that a lot of our customers are focusing heavily on the topic of financial crime, can you tell me about the impact that financial crime is having on financial institutions globally?

Sure. A study by Refinitiv from 2018 showed that over a twelve month period, global financial institutions have spent just under 1.3 trillion dollars fighting financial crime. When we think of financial crime, we think about bribery and corruption; money laundering – a huge challenge and concern; cyber crime – rapidly growing and becoming state sponsored with super sophisticated capabilities to harm firms; fraud – not just a regulatory concern, it hits the bottom line of a bank; theft – which is prevalent in many industries; and then the most painful aspect of financial crime is that it finances and drives human trafficking, as mentioned in a recent blog authored by my colleague Paul Lashmet.

Refinitiv has also estimated that the losses over the same period is 1.45 trillion dollars. So, as an industry, if you look at the numbers, we are not doing a very good job. That was my first reaction. We are spending a lot of money and we are only catching one percent of the criminal proceeds that authorities are able to capture – based on recent EU-wide data.  From this one would assume that while the problem is highly complex, that despite the improvement we have a lot of room for improvement. The secret of this now is that we need to become much more effective and innovative on where we are spending this 1.3 trillion dollars to improve what we are doing.

Earlier you mentioned that digitisation is a fast growing trend in financial services, is this where the gap lies and how criminals are infiltrating institutions?

Absolutely. Everything that we talk about on the topic of digital transformation – whether its on the customer side or risk or elsewhere – provides potential new openings for criminals to exploit and commit sophisticated financial crimes. 

Our recent eBook has highlighted many challenges in the area of financial crime. Customers today, particularly the millennial generation, want access to everything On-Demand. They want real-time, personalised interactions. All the institutions are focused on generating real-time streaming capabilities. The investors expect that the firms are protecting the customer as well as their own business. For example, if an institution is hacked in the money laundering space, it has a huge impact on the brand credibility of a bank. 

The other aspect is open banking. Open banking is really API hooks, PSD2 and other things that are driving collaboration across entities outside the bank. This enables new capabilities and new ways to meet the needs of customers.  This transformative effort is a great opportunity for the banks and fintechs, but at the same time it potentially opens up new avenues for criminals. These criminals are able to leverage the fintech innovations and piggy-back on them for things like deposit fraud or creating synthetic IDs. These criminals are super sophisticated and it’s way beyond the days of counterfeit cheques. Some of the things available on the dark web are shocking – you can buy a passport on the dark web for around 1,000 dollars. Ultimately the Internet and digital technology has created amazing advances to human life but has also enabled criminal networks to spread across multiple jurisdictions, which makes it extremely difficult to identify and stop erroneous behaviour.

And presumably some of the best hackers, data scientists and technologists are working in financial crime?

Hopefully not, but I think you are correct! Unfortunately these people are incredibly skilled and, looking outside of criminal networks, financial crime can also be state sponsored. Their sophistication in this space are almost certainly greater than criminals, but their purpose is slightly different as they attempt to disrupt countries and economies. 

What have financial institutions been doing to tackle the challenge of financial crime to date?

Remember that a lot of this is driven by regulation. It varies by country, but generally regulators have to approve what is being done by institutions on most areas of financial crime. This is done to ensure there is sufficient protection on consumers, but can sometimes delay innovation. In general, regulators are very cautious in the area of machine learning and AI. They are very reluctant to allow financial institutions to use “black box” machine learning technologies because it doesn’t allow consumers to understand why decisions are being made, for example, when they apply for credit or loans. 

In the US, regulators are now encouraging banks to use whatever advanced “black box” machine learning techniques they can muster to combat financial crime. So, if you’re a data scientist, financial crime is the best place to be involved in analytics because your being encouraged to experiment as long as you use responsible tools and techniques. The regulators are rewarding institutions who demonstrate innovative ways to fight financial crime responsibly. If, through an innovation pilot program, gaps in a current compliance program are exposed, supervisory action will not necessarily be taken. They are even setting up pilot project offices which are sharing information directly with institutions to expedite the approval process for getting these innovations into production while sharing best practice experiences with the wider financial services community.

Can you explain the role of data within the context of financial crime and how Cloudera can help to address gaps in the current approach?

If you think about the different areas of financial crime – anti-money laundering, fraud prevention, cyber security and internal surveillance – they are completely different functions. Different people, different skills, different regulators, different reporting requirements and so on. You don’t necessarily want to attempt to integrate all of this from an organisational perspective.

What we are trying to emphasise from an efficiency & innovation perspective is that you can add an enterprise data layer – a financial crime data lake. You want an environment where all of this data resides with the same quality control, metadata management, data security and consistent data governance. This will give you an ability to leverage models in different areas to improve anomaly detection and other factors that are somewhat common across the various financial crime areas.  For example, a customer can use models from anti-money laundering data and apply them to signals in fraud to see if you get unique fraud identification signals. 

You could enhance your Know Your Customer (KYC) data so you have a richer database which can help to improve internal surveillance for other activities. This also provides a significant opportunity to improve the Customer360 capabilities on the revenue side of the business from the investment in enhanced KYC capabilities.   This is why the KYC topic ideally needs to be viewed from an enterprise perspective not just a financial crime requirement.

There are all kinds of data sources that need to be taken into account for financial crime. Cloudera provides an enterprise data cloud that gives financial institutions an agile & scalable  platform to manage data and build very sophisticated machine learning capabilities. We don’t provide applications, so we rely on our partner ecosystem and our open source capabilities to support innovation programs.  Our 100% open source approach and partner ecosystem enables our customers to build out a holistic financial crime data lake. We work well with legacy financial crime application providers as well as newer niche of innovative fintech vendors in this space. This means our customers can modernise their approach at the data layer instead of ripping and replacing all of their existing solutions and applications.

Are there any customers that are working with us in the area of financial crime and any examples that you can share?

Yes, we have a lot of examples, and fortunately some that I have permission to share! Mastercard have built an anti-fraud solution on top of our platform, United Overseas Bank use the platform on anti-money laundering, and Western Union have built a machine learning solution for fraud prevention. There is a lot more detail on all of these use cases that you can find in the FinCrime eBook that we launched recently. .

Thanks Richard! As you have mentioned, we have some other assets on the topic of financial crime. If any of you reading this article would like to see more, please follow the links below:

Ross O'Neill
Ross O'Neill

Senior ABM Manager, EMEA

Leave a comment

Your email address will not be published. Links are not permitted in comments.