We spoke to George Dunning, co-founder of Bud, to get his expert insights

We’re deep in the midst of the fourth industrial revolution, and data is at the centre of it. But with so much data in existence, how can it be properly used? 

This is a problem Bud is trying to tackle using AI; our reporter Robert Welbourn sat down with George Dunning, Bud’s co-founder and chief operating officer, to talk all things data and AI. 

Hi George! Please could you give me a little introduction. 

I’m George Dunning, the co-founder and chief operating officer of Bud. I was the Chief Technology Officer when we started the business; I come from a software engineering background. I did that for the first four and a half years and then moved roles. 

I like to fix things, not just technically but organisationally as well. Can we run things any better internally, can we execute any harder?  I’ve always taken part not just in the building but also the delivery of our technology, as well as making sure the clients are happy and things are running smoothly. I’m highly focused on delivery and execution. 

Wanting to fix things: is that where Bud sprung from? 

Ed (Maslaveckas, Bud co-founder & chief executive officer) and I were new to finance when we started the business; we’re certainly not new now! We’re nine years in, so very well versed in it. Since the beginning, we’ve seen vast potential in transaction data. 

By using the transaction data of a customer, you can understand who that person is and the needs that they have. As we dug in further and further, we realised what an unexplored territory it was, that it was underserved technically. And so we started getting obsessed with it and trying to understand how you could effectively and efficiently – as well as accurately – understand a customer’s transaction data. 

And once you understand it, what’s the value that you can derive from the data? Both for the customer themself, by helping them understand their finances and what they should be doing with them, but also the businesses involved. So lenders, for example, can they use the data to make credit decisions, or to understand their customer base better? 

We’ve just evolved it further and further from that starting point.  

There have been some tailwinds from technological developments, such as open banking making accessing transaction data much easier, and the AI revolution. 

As AI has developed and matured, our stack fits very neatly into it. We’ve been using large language models (LLMs) for seven or eight years as a way to enrich transaction data, but those are very specialised. When you run analytics and understand customer data, you can actually start to interact with generative AI to help it understand the person’s financial data as well. 

You must be welcomed with open arms by a lot of the larger banks who have mountains and mountains of data that they’re dying to understand but haven’t the capacity. 

Being transparent is very dependent on the bank and where their strategy sits. You also have to make sure that your technology fits into that strategy and you’re not offering something that a team is building themselves. Navigating that is always an interesting pattern. 

Banks want to make better use of their data; they’re becoming much more aware of just how valuable it is. Our Drive product does full portfolio analysis instantaneously, combined with the chatbot capability of being able to ask the AI about a customer based on their transaction data. We’re seeing a lot of demand for that. 

Marketing departments are also understanding that this tech will allow them to send out truly personalised messaging. Currently they don’t really have the capability to do that beyond the behavioural information that they have on the customer. We can get it from a transaction level and be able to say, “Look, we’ve seen you spend this much here, we could give you cash back rewards on that.” It allows us to create truly personalised experiences.  

Are you finding banks are quite far into the GenAI journey? 

You see a spectrum for sure. There’s a lot of work to be done. Everybody sees that there’s value, but I think the level of value that people perceive varies. There’s a lot of – as there always is with large scale banks – infrastructure work that needs to be done, not just on tapping the data, but other features too. 

A lot of banks have older, more legacy support systems, including chatbots that are rules-based. If they want to move to a GenAI experience they have to migrate away from these systems whilst making sure that queries are answered in similar ways. There’s a lot of technical migration work that takes priority. 

I do think most of the banks that we speak to definitely have an ambition to make the most of data and AI.They want to create better digital experiences where customers feel like they’re talking to a bank manager again, someone or something that actually understands your finances. 

There’s a lot of work to be done to actually make that come to life, and that’s where we fit in. I think the risk appetite of banks really does vary in how quickly they want to adopt GenAI and what avenues they want to take. People are concentrating on the low hanging fruit first and foremost, which I think makes sense. But as things evolve, people are going to want to see how GenAI interacts with finance and gives the customer true value off the back of this. 

I do think, from some of the banks that we’ve spoken to, how challenging that will be is underplayed. A lot of organisations don’t have their data in a good place; throwing bad data at a GenAI chatbot is never going to result in good outcomes.  

People are really beginning to understand that you need your data, your foundations, in a really solid place. Then the analysis done on the customers’ data, before layering it into a generic LLM, can result in really amazing value. Just giving a billion transactions to a generic LLM is going to end with organisations spending a lot of money and not actually finding a lot of value. 

Are you seeing banks hesitate because they’re worried their consumers might be concerned that their transactions are being rooted through? 

We lock transaction data down so it’s only visible to the bank and the consumer and it’s not being ported out. Nothing is going anywhere outside of the banking environment, so the customer can be completely confident that there’s nobody else snooping around their transaction data. And I think that’s really important, I think people should know exactly where their information is going. 

If I were to do a survey, I expect banks would be more worried about privacy than the users. Frankly, there are some users that are worried, but people tend to be driven by the value associated with a thing. 

It’s about positioning, isn’t it? If you say it to 100 consumers, “We’ve looked through five years of transactions to bring you this offer,” they might be put off, whereas if you say to them, “You can get a lower mortgage rate,” it’s a no-brainer. 

You can already go one step between and say, “Look, we’ve seen your mortgage payments have gone up, would you like us to explore a new mortgage rate for you?” People do want that personalised approach but you’re absolutely right, there’s a balance there. It’s about not being intrusive. Customers aren’t interested in, “Hey, we’re consistently analysing the last five years of your life.” 

That being said, I do think, especially with the likes of Monzo, Starling and Revolut, people expect their bank to be utilising their transaction data to give them value. 

I agree, I’m a Monzo bank user and I love their offerings. But there’s always the other side of it, there’s nothing more depressing than my bank saying to me, “You spent X amount of money on ice cream this month.” I don’t need to be reminded of that! 

Going back to Bud specifically; you’ve started your European expansion and opened your new office in Lithuania. How’s that going? 

It’s going great! We’ve had people in Lithuania for a little while; this announcement was off the back of finally gaining the regulation that we needed to move into the region fully. Europe’s always been on our doorstep, and we’ll steadily expand as we bring our data intelligence services to new markets. 

This just gives us the flexibility to do that more quickly and react to client demands in a way that previously might have been a little bit more challenging. The UK is obviously still one of our primary markets; we’ve also made huge inroads into the US and have expanded our client base dramatically over there. We’re seeing heavy demand for our products and it’s extremely exciting for us.

Image: Bud

Robert Welbourn
Robert Welbourn is an experienced financial writer. He has worked for a number of high street banks and trading platforms. He's also a published author and freelance writer and editor.