No Shortcuts or Detours. How a Team of London Engineers Is Tackling The Fraud in the On-Demand Economy

Photo of Tomasz Grynkiewicz

Tomasz Grynkiewicz

Updated Feb 27, 2024 • 8 min read
Untitled design (7)-1

For a casual observer, the fintech scene looks like a uniform rock: it is all about digital banking alternatives (such as N26, Revolut, Tandem, Yoyo Wallet or Monzo).

As in any industry, the same logic applies here - companies with a B2C business model attract customers the fastest, raise the largest investment rounds and, in the end, grab the most headlines.

In fact, the landscape is full of innovative companies working behind the scenes to solve varied challenges. This is why I chose to speak with Nick Lally, co-founder and Chief Operating Officer of Ravelin, a machine learning fraud detection firm.


Ravelin was founded in 2014 by former engineers from a taxi-hailing app, who observed the impact of fraudulent payments on the global on-demand economy. The challenge they are trying to tackle is significant - online fraud is outpacing the volume of e-commerce sales (source: LexisNexis Risk Solutions’ report, “2017 True Cost of Fraud”). The industry puts the so-called ‘acceptable’ level of fraud at 1 percent of total sales. Nonetheless, this is costly - LexisNexis estimates that for every dollar of fraud mid to large e-commerce companies incur $3.37 in costs, which includes chargeback fees, merchandise replacement, and employee costs.

With new sales channels and on-demand apps, the fraud level is ever growing. That is a “godsend for fraudsters”, as Martin Sweeney, co-founder of Ravelin, put it. “The instant gratification element of two taps is seen in other areas of retail, with same-day delivery and click-and-collect. That’s a godsend for fraudsters because they rely on the consumer having no time to notice anything’s wrong” - he said in an interview.

Stats confirm the growing problem. “Mobile payment fraud already accounts for 2.6 percent of total revenues for merchants who offer that capability,” said Paul Bjerke, vice president, fraud and identity management strategy, LexisNexis Risk Solutions.

Ravelin aims to lower the overall fraud level to 0.1 percent. The company co-founded by a team of five (Martin Sweeney, Nick Lally, Leonard Austin, Mairtin O'Riada and Stephen Whitworth) offers its solutions to e-commerce merchants and basically any company that processes payments. Ravelin scores every action of every user and processes thousands of signals, from on-site and in-app behaviour to the device used, in order to spot anomalies. It tries to prevent fraudulent payments and account takeovers. As of today, the company serves brands such as Deliveroo (food delivery), Kinguin (video game trading platform), and MyTaxi in detecting fraud.

Read on to learn more insights from Nick Lally, COO of Ravelin. Nick Lally is responsible for the commercial delivery of the Ravelin organisation, and loves building teams and empowering them to succeed.


Nick Lally is also a keen mentor and active in the UK start-up community.

Tomasz Grynkiewicz: What will Ravelin look like in 2020? What’s your primary goal?

Nick Lally, COO of Ravelin: With global initiatives, such as open banking and a bunch of payment regulation changes in Europe, the way e-commerce businesses collect revenue from their customers is going to start to look different in 2020. Ravelin’s role will remain the same; we’ll focus on how we can help these businesses allow the maximum number of good transactions - however they are generated - without falling victim to fraud.

While we do this primarily for payment cards today, and that will be the same in 2020, we will secure a much broader range of payment types, as well as help our customers tackle the growing issues around identity fake accounts, account takeovers and a plethora of terms of service abuses.

What are the major challenges in the industry that could impact your vision?

There is a lot of complexity on how payments are going to develop, and a lot of confusion around where the liability for fraud will fall. It’s a significant education effort for a business such as Ravelin to undertake, so we need to find the channels, bodies and partners who can help us to spread the word.

I’ve read that Ravelin, like many companies with sophisticated machine learning systems, needs to deal with the so-called adoption challenge.

It is important to underline that our machine learning models aren’t there to eliminate a fraud team. They’re there to enhance them. Our platform gives a business’ analysts the tools and insights they need to win the fight against chargebacks and fraud.

A lot of startups fail, but Ravelin succeeded. What values and principles were crucial in accelerating the growth of the company?

I think honesty is at the core of it. We have probably made more hard calls than most start-ups, and walked away from some deals that appear attractive but would be a major detour for us. Instead, we’ve focused on delivering honest, straightforward advice and expectations to our clients and prospects about what we can deliver for their business. As a fraud detection company, we use multiple technologies, including machine learning and behavioral analytics, to give the best results for our clients.

What are the most important features and values that made Ravelin stand out from the fintech crowd?

Pragmatism is at the heart of everything we do. Ravelin was born out of a merchant environment dissatisfied with the incumbent fraud tools in the industry. We created Ravelin to align with the dynamism of today’s fraud and to constantly adapt to the emerging threatscape.

What are your major markets to conquer?

We started very squarely at the centre of what we know well, which is the on-demand app space. Some of the team at Ravelin worked at an on-demand app previously, and it was the experience gained there that made us realise that there was a gap for a real-time, machine learning fraud tool that combined speed with accuracy.

Luckily these requirements are increasingly shared in other industries, so we’ve started to gain clients in ticketing, travel, gaming and retail e-commerce. As the fraudsters evolve, our business needs to evolve with them. Over the next few years, we expect to see some other sectors soon.


What about geographical expansion? European fintech companies are starting to challenge overseas markets. What is Ravelin’s approach?

We operate across 93 countries already, because this is where the customers of our clients are placing orders. From Lima to Liverpool and from Dubai to New York. We serve these clients from our London office currently, and we are likely to grow that soon, but I can’t say anything more!

These are huge opportunities, but the race is to understand what is going to be useful to the market, understand how to productize it, and then sell it. No one wants to understand these regulations, they just want to comply and get on with their lives. The winners will be the companies that can abstract that for businesses and allow them to successfully grow their businesses and revenues in this new era.

From your point of view, what is inevitable when you start a fintech company in 2018? What would be your advice to other players?

Solve a real problem and solve it really well. By definition, that problem will need to be quite small initially. And be patient. Part of success is hanging around long enough.

Ravelin is a great example of machine learning. However, if you were to tip your hat to any company you find inspirational, which one would you choose?

Monzo. Great team, great product. I use it every day.

Photo of Tomasz Grynkiewicz

More posts by this author

Tomasz Grynkiewicz

Team Leader
How to build products fast?  We've just answered the question in our Digital Acceleration Editorial  Sign up to get access

We're Netguru!

At Netguru we specialize in designing, building, shipping and scaling beautiful, usable products with blazing-fast efficiency
Let's talk business!

Trusted by: