Monzo: how the bank of the future uses AI

Credit: WanderlustChloe.com

Monzo is changing perceptions of financial institutions, one pink card at a time. We speak to their machine learning lead on how AI is transforming banking.

When Apple launched the catchphrase “There’s an app for that”, perhaps not even they could predict the industries that would be turned on their head by smartphone technology.

Uber has revolutionised the way we think about taxis, to the point where apps for takeaway like Deliveroo, or Slack for communication, have been referred to as “The Uber of…” another specific industry. However, perhaps the biggest success story in recent years of app-based businesses is Monzo, a digital-only bank famous for its coral-pink cards.

Monzo stands out from the crowd and not just for its card shade. It’s a given that major financial institutions should offer their services online these days; that high street banks should come with an app for checking your balance. Monzo takes that a step further, leading the charge on so many banking features that are being copied by others. Auto-budgeting was a big favourite when the app launched; since then, the option to round up your change and leave it in a savings pot has become increasingly popular too.

Finance is changing and innovation runs through the heart of Monzo. The company has caught the imagination mainly from Millenials because of its willingness to do things differently. Monzo uses machine learning to improve its service and understand its customers a little more.

It’s perhaps an unlikely success story: a bank that has become fashionable, completely without branches. “I’ve chatted with data scientists across the industry who are in some of the older banks,” says Neal Lathia, Machine Learning Lead at Monzo. “To a certain extent, all of them are looking into AI.”

What makes Monzo different?

“I was first introduced to machine learning when I started a PhD at University College London and became interested in how recommender systems work,” says Neal. “At the time, I was using tools like Last.fm to discover music and Netflix had just announced a million-dollar competition to improve their recommendations. From there, my career has always focused on machine learning and data science.”

Neal did years of post-doctoral research. From there he was involved with a health tech start-up before consulting for a few companies and joining Skyscanner. He was introduced to Monzo as a customer by his sister, who signed up in what he calls, “The old pre-paid card days,” before the company had a full banking license.


The technology itself has matured so much in the last few years that AI is now where mobile apps were a decade or so ago: sort of new, but emerging quickly.

Neal Lathia, Machine Learning Lead, Monzo

“That old version of Monzo didn’t have savings pots, loans, shared tabs, or many of the other brilliant features that make Monzo unique today,” he says. “But I was excited by the potential of what it could become, and specifically excited by the prospect of being able to build tools that leverage rich financial datasets to help make money work for millions of people.”

Neal joined Monzo in February 2018 as a Senior Data Scientist at the business. “I don’t count myself as an expert in all of the new banks that are out there today,” he admits. “Most of them are saying similar things – like putting customers first and being transparent. Someone who joined Monzo recently and was speaking about this at a company-wide meeting made a comment about this that rang true for me: Monzo isn’t just speaking publicly about these values, but also living by them.”

Machine learning is at the core of these values. Neal says that the company are considering “stepping outside” of the applications associated with traditional banking. One idea blogged about recently is trending spending, a feature which would allow customers to discover places in their location based on financial transactions. Such data would presumably require AI to implement.

“We are going to work on building up the foundation for machine learning in the core areas of Monzo,” Neal says, “like we are doing with customer service right now.”

Monzo has already introduced AI to customer service

“We have a very small team of machine learning scientists at Monzo: there are four of us total,” says Neal. “Right now, our major focus is on customer service.”

With customer service being a key area of concern for so many banks, Monzo is taking a different approach. There are two complementary sides to how they use machine learning in this domain.

“First, we partner with product teams to improve the help sections of the app: for example, we run the systems that allow customers to search our help articles,” Neal explains.

Neal Lathia of Monzo. / Credit: NS Banking

“When customers go to the chat screen and get in touch with our customer support, we turn towards the second side: building tools to help our customer support agents to help customers faster. For example, we have a system that recommends to agents how they can respond to different customer queries.

Neal believes that this technology will become the norm in the coming years. Whilst many banks are looking into how to implement AI, very few are already using it to benefit their customers.

“The technology itself has matured so much in the last few years that AI is now where mobile apps were a decade or so ago: sort of new, but emerging quickly,” Neal adds.

What does the future hold for Monzo?

“Shortly after I joined Monzo about 18 months ago, we celebrated a fun milestone: we had just surpassed half a million customers,” says Neal. “We now have more than 2.5 million UK customers and are launching in the US while continuing to release new products – like loans – here in the UK.”

The sharp rise is not predicted to slow any time soon. Over a year ago now, the company documented how it planned to accumulate a billion customers: that’s a lot of customer service, for a start.

Artificial intelligence will no doubt continue to be central for Monzo in learning about their customers and expanding accordingly.

“I see machine learning as one of the many tools that we will use to help customers discover, understand and make decisions about their financial lives.”

Luke Conrad

Technology & Marketing Enthusiast

Birmingham Unveils the UK’s Best Emerging HealthTech Advances

Kosta Mavroulakis • 03rd April 2025

The National HealthTech Series hosted its latest event in Birmingham this month, showcasing innovative startups driving advanced health technology, including AI-assisted diagnostics, wearable devices and revolutionary educational tools for healthcare professionals. Health stakeholders drawn from the NHS, universities, industry and front-line patient care met with new and emerging businesses to define the future trajectory of...

Why DEIB is Imperative to Tech’s Future

Hadas Almog from AppsFlyer • 17th March 2025

We’ve been seeing Diversity, Equity, Inclusion, and Belonging (DEIB) initiatives being cut time and time again throughout the tech industry. DEIB dedicated roles have been eliminated, employee resource groups have lost funding, and initiatives once considered crucial have been deprioritised in favour of “more immediate business needs.” The justification for these cuts is often the...

The need to eradicate platform dependence

Sue Azari • 10th March 2025

The advertising industry is undergoing a seismic shift. Connected TV (CTV), Retail Media Networks (RMNs), and omnichannel strategies are rapidly redefining how brands engage with consumers. As digital privacy regulations evolve and platform dynamics shift, advertisers must recognise a fundamental truth. You cannot build a sustainable business on borrowed ground. The recent uncertainty surrounding TikTok...

The need to clean data for effective insight

David Sheldrake • 05th March 2025

There is more data today than ever before. In fact, the total amount of data created, captured, copied, and consumed globally has now reached an incredible 149 zettabytes. The growth of the big mountain is not expected to slow down, either, with it expected to reach almost 400 zettabytes within the next three years. Whilst...

What can be done to democratize VDI?

Dennis Damen • 05th March 2025

Virtual Desktop Infrastructure (VDI) offers businesses enhanced security, scalability, and compliance, yet it remains a niche technology. One of the biggest barriers to widespread adoption is a severe talent gap. Many IT professionals lack hands-on VDI experience, as their careers begin with physical machines and increasingly shift toward cloud-based services. This shortage has created a...

Tech and Business Outlook: US Confident, European Sentiment Mixed

Viva Technology • 11th February 2025

The VivaTech Confidence Barometer, now in its second edition, reveals strong confidence among tech executives regarding the impact of emerging technologies on business competitiveness, particularly AI, which is expected to have the most significant impact in the near future. Surveying tech leaders from Europe and North America, 81% recognize their companies as competitive internationally, with...