Everything is Fintech Live: Caywood's Founder & CEO, Megan Caywood

0:00Oliver Smith: Part of the challenge is that banks were getting a lot of revenue from inertia. Customers wouldn't switch. And so that friction was part of it. It's like it was so painful to do anything. It's such a common joke now, but people would get divorced more often than they would switch a bank account.

0:23Megan Caywood: Hello and welcome to Everything is Fintech. My name is Oliver Smith, Head of Content at Money 2020 Europe. I'm delighted to be joined by Megan Kaywood, CEO and founder of Kaywood, and previously Chief Product Officer at Barclays, and before that, Chief Platform Officer at Starling Bank. There we go. There we go.

0:43Oliver Smith: You got all the CPU.

0:44Megan Caywood: Thanks for joining us, Megan. You've had a really interesting journey through your career. I guess I'd like to start just identifying the different pieces of it. Maybe first of all, could you tell us a little bit about the work you did at Barclays, I think, firstly?

0:59Oliver Smith: Yeah, absolutely. So when I joined Barclays, it was kind of at the height of digital transformation. So the challengers had come to market, predominantly in the UK, where we had new regulation that enabled us to come to market, get the full banking license, build a bank from scratch. There's a lot of innovation happening there. And they brought me in from Starling to oversee two things. So one was as head of digital strategy to start with, it was across all consumer and that was US and UK. So you can think of that as payments, mortgages, ops and tech, retail, SMB. So it was kind of the whole nine yards of consumer banking.

1:33Oliver Smith: And so I focused on that. And then the second part was a real deep dive on in terms of core banking. So looking at the existing infrastructure we had, how we wanted to evolve that under our digital transformation strategy. And then once we had that ratified by the board, I stepped into the chief product officer role to lead the execution of that. So we did an internal build called the Next Gen Consumer Banking Platform. So, yeah. So, yeah, I built a new consumer bank. It's what now powers the Amazon co-brand card that they offer as well as their internal products.

2:04Megan Caywood: And tell us a little bit about, you know, working in an institution like that. It's not always easy to innovate in a large bank, in a large organization. What was that journey like?

2:13Oliver Smith: Yeah. So when they had originally reached out, the leadership team at Barclays, I was so stuck in it, Starling, and just loving it. I really didn't want to consider moving anywhere. But they asked me to come and network with the leadership team and get to know them. And what I found in that process was that the C-suite of Barclays was very switched on to what digital enabled for consumer outcomes. And they really wanted to bring that magic into Barclays. And so I felt like we had a shared vision from the top of the organization. So I was like, if I have that backing, that's a sufficient framework to then try to move the rest of the 80 ,000 company in a 330-year-old organization, which is highly confident in hindsight.

2:50Oliver Smith: But it was really good. I think that the thing about Barclays is that it was unique in that so many people were willing to change. I think a lot of people really wanted to both drive shareholder value as well as good consumer outcomes. And I think that made it quite unique. They were always kind of leading the way in terms of consumer innovation. And so I think as a whole, I think they culturally had a willingness to change and do things differently that worked well for them. That being said, being such a big organization, what I saw is it could be very different across different groups. They had different incentives and goals and remuneration for their particular bonuses.

3:27Oliver Smith: And one group might be tied to reducing tech debt and another one new product innovation. So kind of like aligning across the bank as to what's our shared vision? What do we need to drive forward first? Where are we going to get the most in terms of both consumer outcomes as well as, of course, reducing cost base as well as, you know, reorganizing to work as a tech company? kind of hitting all of those at once was probably the challenging thing. And honestly, one of the first things that we looked at was just, you know, if we were to shift under a digital transformation strategy to be, you know, moving from legacy bank to operating like a tech company, what does that look like?

4:01Oliver Smith: And the reality was, is yes, there's a technology strategy in terms of, you know, having a scalable microservices cloud-based architecture that that's part of it. But more fundamentally was the way that we worked. Yeah. And so evolving our operating model was just as important to us as evolving our tech stack. And so, you know, figuring out how to create, you know, small lean teams who had a mind of being able to experiment and, you know, drive customer outcomes and really own the outcome and work cross-functionally, making sure all the cross-functional leaders were in the same team so that they can make decisions quickly and easily.

4:34Oliver Smith: All of those things were kind of part of the journey. So it was a big cultural shift, both in terms of the tech and the ways of working. but I think yeah all in all it was a really good journey with how the leadership supported and how the team kind of incrementally moved towards it.

4:48Megan Caywood: Yeah and what years were that was that 2020?

4:51Oliver Smith: 2019. 2019 to 2024.

4:54Megan Caywood: See in my mind those are the years where traditional banks you know the disruption that had been caused in the 2010s by Starling and Monzo and others we will speak about I think it felt from the outside like the big banks had woken up and really were playing the kind of catch-up game of like mobile banking. We need to be way better than this. Our app can't just be rubbish anymore. It has to be really good. And I think that period that you were there felt like it was that catch-up time. All of those features that we took for granted in fintech land, whether it was check scanning, whether it was freezing your card, all of a sudden the high street mainstream banks were picking those up.

5:33Megan Caywood: Yeah, absolutely. But I want to switch now to what you did before Barclays, because Starling Bank in the UK, for those who aren't aware of it, was really one of, or if not the first, proper digital challenger bank that came to market and redefined what it meant to be a neobank or a challenger bank. Tell us a little bit about your time there.

5:53Oliver Smith: Yeah, absolutely. So I had moved over from San Francisco after meeting Anne Bowden to join as chief platform officer in building this new challenger bank. And like you said, we were the first digital challenger in that we got the full banking license, we built the bank from scratch, and we were really looking to innovate on the user experience. There's a lot who were coming to market at that time, but it's prepaid cards. So they very much saw the opportunity in terms of the user experience. But actually innovating on the tech stack and the business model with a full banking license was something that we were doing quite uniquely.

6:21Oliver Smith: And effectively, what we saw was that there had been so much innovation in the world of consumer tech, particularly with the advent of the iPhone, that now what consumers were expecting in terms of how intuitive and beautifully designed a service could be was so different from what it was even 10 years prior. And that hadn't yet come into the world of banking. Within the world of banking post-financial crisis, we were so focused on the new fines being levied and the new rules and regulations that this gap just naturally emerged where at Starling we were able to really come in to differentiate.

6:49Oliver Smith: So we had these two key elements to it. So one, we didn't launch until we had the full banking license. And uniquely is we didn't build on a third-party platform. So we built out the back-end tech. So we consider ourselves a full-stack bank. And what we really wanted to focus on was taking all of the friction out of banking. So make it super easy to open accounts. You open a bank account in less than three minutes on your phone, provision a card to Apple Pay, can add money to it instantly and be spending in five minutes. and then getting the real-time push notifications for each transaction, having those go into insights into spending, integrating things like the Google Maps API so you could see the location of the transaction.

7:26Oliver Smith: All things I think we probably take for granted now. But if you rewind to 2016, that was quite exciting. You know, the ability to freeze a card was like quite novel, even though now it's like the most common feature, I think, in digital banking. But yes, we were really wanting to be leaders in terms of making banking easy and helping customers make better financial decisions.

7:45Megan Caywood: I think, as you made that point, I think we forget what traditional banking was like in the mid-2010s. We forget just how bad it was. And the apps, the digital experience was so second class, opening an account being a great example. You know, paper forms, going to the high street store, like none of what we now take absolutely for granted.

8:06Oliver Smith: Oh my gosh, I used to joke with people that when I first moved to London, like, because it's funny being in the US, because it was not good in the US, but it was terrible in the UK. Okay. Like I remember I had, I won't even name the banks. I don't want to shame them too much, but like I opened a bank account, which was a horrific experience, as you say, especially as an expat. But then to log into the mobile app, you had to generate a digital secure key, go online, remember your first employer, your mother's main name, and then it would give you a code that you go back into the app to then log in.

8:31Oliver Smith: And it was just like, just the basics of opening the app were so hard. And it felt like there was a fear around security and everyone's question like, is, you know, what's secure and how do we do it? And rather than actually thinking it through, they just put all the burden back onto the customer. So everything was high friction. Everything was painful. And then the CMA, the Competition and Markets Authority in the UK, they did an analysis and they found that part of the challenge is that banks were getting a lot of revenue from inertia. Customers wouldn't switch. And so that friction was part of it.

9:02Oliver Smith: It was so painful to do anything. It's such a common joke now, but people would get divorced more often than they would switch a bank account. But it was partially because the account opening process was difficult. Comparing fees was difficult. Everything was difficult.

9:15Megan Caywood: So I think if you spoke to, you know, someone from Gen Z or Gen Z or whatever today and told them that journey to like log in to your account with all of those codes and various things, they would think you were making it up. It can't have been that difficult, but it was.

9:29Oliver Smith: Oh, I know. My pages are so fast now. Like, back in my day. I had to go in and fill out forms.

9:35Megan Caywood: Megan, I want to bring us up to speed with where you are today. So you left Barclays and you've gone out independently. Tell us a little bit about Kaywood and what you're doing now. Yeah.

9:45Oliver Smith: Oh, this is the perfect segue because I think where we are now, like within like a decade, we'll be regaling like, I don't know, whatever it is, Gen Alpha, about what, you know, banking was like back in 2025. Like you actually had to open an account and manage it yourself. Like I think the next wave is we've had digital transformation over the past, you know, 20 years probably, but probably 10 years within finance. But the next wave is agentic transformation. And agentic AI are systems that don't just predict, but they act within regulated frameworks, they execute, they audit, and they learn autonomously.

10:17Oliver Smith: And what that means for the world of personal finance is that you're no longer going to have to worry about having a good user experience or a pretty dashboard or better insights, because you'll be able to have a CFO in your pocket who acts effectively. So that was the first premise I really came to with KWIT. It's like AI is intelligence. The intelligence that we can now bring into financial services is phenomenal. This was like when I started my career in San Francisco and into it. That's where Scott Cook, the founder, started was with Quicken, like sitting at a dining room table with his wife, like with all these, you know, different bills and things everywhere.

10:48Oliver Smith: and they both have like a Harvard MBA and they look at each other like, gosh, if it's this hard for us to manage our finances, how's the average American like, you know, doing on a daily basis with this stuff? It's complex. So they created Quicken, which obviously Intuit's a phenomenally successful business so that worked out well for them. But what we have now is whenever we have a new technology come to market, it shifts the art of the possible in terms of the experience. It doesn't necessarily always create new industries. It's like the classic examples of like Uber, Netflix, like the taxi industry existed, You know, streaming videos, you know, existed, but it's the way that they delivered it with new technology created a totally new experience and a totally new business model.

11:27Oliver Smith: And my vision is that AI will create a totally new experience and a totally new business model for the world of finance. And part of that is because it's Starling and at Barclays, what we were trying to do is take everything that currently existed and just make it easier. Just automate it as much as you can. Take out as many steps as you can. Just try to make it easy. But effectively, customers still needed to open the account. They need to know which account to open. They still need to manage their finances. And the challenge with that is for a lot of people, they're not experts in finance. They don't want to be.

11:55Oliver Smith: There can be a lot of anxiety around financial management. And that's where the brilliance of AI can come in to help them make better financial decisions, but then execute on their behalf. Because the AI actually executes on the existing rails, but it will move money and it can act autonomously without you having to do everything yourself. So that's the shift where it's like you don't have to worry about having a beautiful user experience because you're not going to be experiencing it. Your agent will. So that's part of the first insight for it. Now, where we've shifted and where we're actually focusing now is on the B2B side, because the insight I had when working on that is having worked in banks within the CFO function.

12:30Oliver Smith: There's still so many parts of finance internally to a bank where agentic AI can be applied to not just automate, but to enable better decision making within banks. So examples that are quite easy is like daily liquidity and compliance. So banks have to have a certain amount of capital thresholds that are met that for, you know, the amount of deposits and lending that they do, they have to maintain certain thresholds to ensure customer deposits and report that to the regulator on a daily basis. That's something that currently happens where they're extracting six different CSV files from like payments and deposits, investments.

13:00Oliver Smith: They're reconciling that in one Excel sheet and then it's going up and it's being approved and it happens every day and it's very manual. But that's something where you can quite easily delegate that to an agent who can act autonomously. They can transparently be able to reconcile each of those different pieces of information and data internally, create a transparent audit log that's useful internally as well as the regulator for the sign off. So there's a lot of opportunities within banks to do that same CFO in your pocket who acts, but effectively internally to banks that we're currently working on to start.

13:28Oliver Smith: And then secondarily, it will be to the consumer, same CFO, same kind of agentic spine for finance. But we're actually starting with a focus on banks and internal financial operations.

13:37Megan Caywood: I want to explore both of those threads separately. So the first one was the, I'll call it the distribution disruption, the kind of way that we interact with financial services may not be direct in the future. It may be indirect. It may be through agents and through agentric. Do you see that as a revolution which is going to be forced upon the banks? Is this going to happen to them? Or is this something that they need to do themselves, that they need to get on board with and open up in a way?

14:09Oliver Smith: What's interesting is even when I was at Barclays, even like five years ago, we were talking about money on autopilot. We've been talking about it for so long, to be honest. It's just the technology wasn't quite there. I think there's actually a brilliant TED talk that it looks at the most successful startups and what makes them successful. And it looks at is it your team, your business model, the amount of capital you raise, the quality of your vision or the timing and the variable that like most correlates with success is just timing in the market. So there's like effectively an Instagram before Instagram that like the technology wasn't quite right to the user experience wasn't like quite there.

14:40Oliver Smith: Similar with like YouTube. And I think that's where this money on autopilot, this idea that we can have the CFO in your pocket who acts, something that the banks have actually been quite keen on and wanting to implement. It's just the technology wasn't quite there. So I think, I mean, not all banks, but I think a lot of them are actually quite excited about Agentec AI. Like even looking at Citi and some of the reports that they're putting out and some of the testing that they've been doing around avatars and things consumer and internal. I think we're all quite excited about agentic AI and want to take advantage of it but I do think it'll probably be like the challenger banks where when you're working with a new technology it's just a bit easier from a startup perspective than doing it in a large regulated bank at scale yeah so it'll probably be like the most innovation from the startup perspective the incumbents will learn from that they'll both kind of shift in the same direction at the same time but I don't know that it'll be forced upon them so much as they'll they'll want to make it happen Sure.

15:32Megan Caywood: When I said forced upon them, what I was kind of referring to was, so OpenAI acquired a fintech recently that did sort of like personal finance management app kind of thing. And when I see what they're doing with agents where, you know, agentic commerce is starting, the very early doors of it is starting where, you know, transactions are happening without the consumer going on the website of the thing. Expanding that into financial services where, you know, potentially the consumer isn't going to the website of the bank to open an account. The agent is doing it on their behalf. And that sort of, you know, that disintermediation of that relationship is interesting to me because I wonder what the banks and what brands feel like about that, you know, if they don't have that direct relationship.

16:19Megan Caywood: Do you think that's going to happen?

16:20Oliver Smith: It is a good question, yeah, because that gets to the question like, will the banks be developing the agents or will everyone have their own individual agent? Yes. Because if it's not the bank's agent, you're right, they're disintermediated and now the dumb pipes are getting dumber and it's the whole scenario where they're still doing the thing of like holding securely deposits. They're still enabling transactions and payments, but they're not holding the customer relationship. So their ability to upsell higher profit financial products and all the rest is obviously going out the window when someone's not interacting with them directly.

16:46Oliver Smith: So I do think that part is very competitive and disruptive. um and yeah it's just tbd like will a bank particularly embrace it and have their own agents will people have their own individual agents i don't think we know yet yeah um but yeah it's a good point in that case it would be forced upon them i think uh it's also accelerating so much faster than digital did like when you look at the adoption rates of open ai it's like faster than any other company in history yeah and so that's where i think it'll probably be within like two years that we have the answer to that question but i think yeah it's very accurate i I think there's a quote from probably a decade ago from Oracle, I think, where it's like people need banking, but they don't need banks.

17:25Oliver Smith: And I think it's just getting truer and truer as time goes by. So it's like, yeah, the retail side of banking is, yeah, always under pressure. Super interesting.

17:33Megan Caywood: Okay, let's switch to the other side of it and the CFO in your pocket, both within banks. So within banks using AI and Argentic, but also the customer piece. Tell me a bit more about the work that you're doing within banks. What are they asking for your help on and where the CFO example is a really good one? What other sort of areas do you see advantages coming?

17:53Oliver Smith: Yeah, absolutely. So I think one of the ways that we can think of AI is it's a general purpose technology. So pretty much every industry, every function probably has like a reasonably good use case for how AI can augment or expand or enhance that particular capability. the round table the other day actually with MasterCard and Microsoft and others was and Deutsche Bank and I should gosh I'll have to name them all now and SoundHound and mostly AI it was fascinating because we were talking about AI and the workforce I think we all agreed that across different functions there's this opportunity to augment in an inter-intelligence partnership for the art of the possible and from my background in banking one of the easy ways that we can do that that I've identified with Agentec AI is within the finance team and it's partially because you want an area that has clean data.

18:36Oliver Smith: I think Greg Amasterhart is mentioning this, like the quality of the data is important for training these agentic systems. And so within the financial team, you have this clean source of data. And then you also want it to be able to have a clear set of guardrails and policies and frameworks that drive the decisions, because you want to be able to log every decision for how the agent is working and put that into a human read journal that can be audited by the regulator. So post-financial crisis in the UK, the regulator created the senior manager's function. And within that SMF, those who are taking certain decisions and actions within the leadership or material risk-taking part of the bank, they have to be qualified for those roles, and they have to be able to evidence how decisions are made.

19:16Oliver Smith: And they can delegate actions but not responsibilities. So it's an area that for an agent, we can have this clear set of kind of rules for the system for how it can operate, and then can translate that into a really clear audit log. So it's an area where we can get regulator approval, do something really clean, show the Glassbox AI, also drive automation efficiency. There was a report by McKinsey and BCG where it's something like$800 billion is tied up within banks, CFO offices in terms of automation efficiency that can be unlocked. And so it's just an area where it's high need, Agentec AI can fix it, and we can do it in a regulator-friendly way.

19:51Megan Caywood: And then is that doable in a sort of a repeatable, because one of the challenges with LLMs, with Argentic is that sometimes you ask the same question and you get different responses each time. Is it doable in a way where it's repeatable and auditable that would appease and would, you know, go along with what the regulator needs and what the rules require?

20:13Oliver Smith: Yeah, that's absolutely, it's part of the magic of it, actually, because what we're building is effectively the decision engine and then the auditability. And as we refine those loops in one bank, it almost creates like a network effects platform because then that loop is something else that another bank can take advantage of. So once you refine that in terms of the decision-making capability and the process loop for how the decision is made and how it's audited, then that creates something that is kind of like viral in terms of the utility of everyone else on the network. So that's, yeah, an excellent point.

20:40Megan Caywood: I've got to ask the skeptical question because a lot of people might be listening thinking, agentic there's a you know there's a lot of talk about a bubble at the moment and whether the excitement and enthusiasm around this technology may suddenly wane and things may go back now in the past we've seen similar excitement that hasn't you know the cloud cloud computing i would say is a great example a lot of excitement and actually translated into a huge change took a long time but it did it did kind of pan out um with agentic we're we're yet to kind of reach the point of like belief that this is the future.

21:15Megan Caywood: Do you have concerns that the agentic AI sort of bubble could pop and the hype could fall out of what's happening?

21:22Oliver Smith: My thoughts on it are, the best way I can describe it is agentic AI is enabled by organizations who empower this interintelligence partnership. Because AI could just be used like an LLM or a tool that provides brilliance upon request. You ask it to create a report, refine an email, do some software development. And it operates very effectively within those functions. But to get to the agentic side, it's very much about like partnership. It has to be enabled to be empowered. So agentic AI, by definition, is only going to reach its kind of full capacity if we kind of embrace it and adopt it. So part of like, will it actually get past the hype cycle is like, can we shift that mindset to the art of the possible?

22:04Oliver Smith: And part of it, what I loved about the roundtable yesterday is often when we have these conversations around like a new technology in the workforce every time people immediately look at like replacement and loss and fear like what jobs are going to disappear and so i love that the roundtable actually shifted to like actually know like what's possible in terms of new jobs and new roles and how this can unlock more creativity and enable us to focus on like higher order work and i think as that mind shift like mindset shifts where we see like the opportunity rather than like fear from it and like embrace it and create these systems that enable logistic ai then it will be adopted but to me honestly the way it feels is like i don't know because i've i've so often worked with new technologies even when mobile came out like when i was into it there was a skepticism like will mobile actually be everyone was still prioritizing desktop yeah and then online and then mobile and mobile was like it was kind of easy to lead on mobile because no one like it wasn't that big of a platform yet but so i remember early career i'm like i think mobile is going to be the future i'm going to get all this mobile first experience because I feel very sure that consumers are going to increasingly use mobile.

23:06Oliver Smith: I remember Mary Meeker had this internet trends report out of Silicon Valley and you could just see it. I'm like, it's going to be the classic J curve. It's going up. And similarly with agentic AI, that's the way I see it. It just makes sense because you need brilliance, but you need to empower it to act, to learn autonomously, to execute autonomously. And that's where you really get an unlock both in terms of human creativity and enabling us to focus on like higher order work within like organizations, but also kind of the best outcomes from the partnership. so I think it's there's already so much excitement around it I think like the testing will happen people increasingly adopt it I think whatever skepticism around hype will kind of naturally fall away as it starts to be used more and more and people naturally see the capability of it um and then it'll become you know whatever ASI super intelligence and then we'll all be retired

23:48Megan Caywood: fine anyways sounds good sounds good my very last question um is the roadblocks to that future you know using the cloud computing as an example it took a long time for for banks to get on board with that they had the legacy challenges of siloed data you know lots of friction internally to getting that change through what do you see as the roadblocks to that agentic future what's going to slow the journey down and what needs to be overcome yeah i think it really depends on like

24:19Oliver Smith: the organizations that we consider i think within banking the first thing is just that from my experience working on both sides it's it's easy to easier to innovate when you're the startup inside a big bank the reality is that there could be a lot of um risk aversion to new technology sometimes because part of i mean particularly within the uk where we have ask math there's like a lot of liability on individuals if they release a new piece of tech that really um has a negative impact on customers you've seen like as part of our training we'd have these horror story use cases of you know platform releases gone wrong and like your bonus clawed back for like seven years and you have to pay it back with like the employer and I it's it's a whole thing so I think from banks my theory is that they'll be a little bit slower to test and adopt and roll it out purely because of that risk aversion around new technology that can exist is the same around cloud and I love how they'd be like we have a private cloud I'm like you have a mainframe I guess on prem the private cloud so um I think um it but it really just depends like no two two incumbent banks are made equal.

25:20Oliver Smith: Again, Citi's been doing a lot of really impressive work, I think. Even Barclays is starting to innovate with it. So I think what we'll see is certain incumbents starting to innovate, test it out, see what will happen. There will still be that risk aversion that slows it down, I think, from the incumbent side. We'll see startups coming out with a lower risk and a smaller customer base who can innovate a bit faster, who'll show the art of the possible. And that will help to overcome it. But I think within the world of finance, It's always just around, you know, how can you safely test and roll out the new tech at scale without, you know, kind of coming up against some of these like risk frameworks that exist today.

25:54Megan Caywood: Fantastic. Megan Kaywood, thank you so much.

25:56Oliver Smith: Yeah, thank you so much, Oliver.

25:57Megan Caywood: Thank you very much. And that is all we have time for today. To our listeners, thank you so much for listening. Please like, comment and subscribe. And you can listen to Everything is Fintech wherever you get your podcasts. Thank you very much and goodbye.

26:14Thank you.