Maulik Sailor (00:10)
All right, so once we're settling, I'll kick off the evening. I'm Maulik, I'm founder of Innovify and Get Future ReadyCommunity. We're looking to build curate a community of builders and operators and experts within at the intersection of fintech, e commerce and AI. So very specific on what we're looking to do.

And today, you know, is our September event. We are doing it on a monthly basis. And today we've got an expert panel, which we are hosting for the first time this year. So it's gonna be an interesting panel that I will take you through in a in a moment. it's a community, as I said, we are building is going to be a curated community where we don't want people who are just generic and they just like you know, floating around. Instead, we want people.

Who are actually building and operating within the intersection of FinTech AI and e-commerce, and someone like people from which we can sell ideas, knowledge, you know, connect with each other and build something great out of it, right? So that's the purpose. And we curate, we we met every single person who is coming and joining Q of events. And we hope that all of you will have at least one interesting person for you to take away today,

Again, if you're not part of the community, please go and join at joinfeatuready dot com. you keep in the note. again a QR code if you want to scan, if you're not then we'll just move on.

Me and Max, we are the host. Any questions, any comments, you know, if you're happy about it or if you are unhappy about something, please let us know so we can improve and do better. so me and Max, Max, raise your hand. Yeah.

So, what are we going to talk about? Right. So, generally we curate the community or the event based on the problem statements that we are seeing in the industry. We are, you know, hearing from the customers and the partners that we serve. And based on that, we create our thought process that okay, maybe this is a problem statement or an inter interesting trend that we are seeing, and we wanna bring experts.

To talk about that and you know set knowledge based on their own individual experiences. So that's what how we curate the event we month. And today we got an amazing panel, which I'm going to introduce in a moment. but before I do that, I need need to say thank you to Isaiah for having this space for us, you know, and hosting us in this office.

you know, finding space is difficult in London, you know, particularly where you can host offer work events. And for Isaiah to to step in and support this community was amazing. Isaiah, why don't you say a few words about you and the and video? Thank you very much. You're live and

I guess most of you know BDO, but if you don't, yes, BBO Professional Service Firm focus on audits, tax advisory. I specialize in innovation incentives, so I sit within the tax team. And I also have a very special focus in financial services and fintech. anything innovation incentives, R<unk>D tax, funding, that's kind of my area of specialty.

I'm very much interested in this community because of the sort of synergy with the financial services and fintech space. normally for quite some time. used to work together. But had and yes, I'm always interested in technology, innovation, that's kind of my bread and butter. looking forward to the discussions and the conversations that we're gonna have tonight.

Thank you all very much. Enjoy the drinks, enjoy the pizza and let's have good conversation. Thank you. you. Now today going to talk about an interesting topic, which is not an obvious trend, but something that we are we are witnessing in the industry. Right?

Over the past few years, right, if you think the whole last fifteen years about the growth of SaaS and and fintech as a whole, the whole premise was like

segmenting out like you know from the old school big banks and big financial institutes. It was all about breaking it down. You know, you have a SAS which does a particular thing like you know your FX or your compliance or your KYC or you know reconciliations or whatever, right? So it was all about a particular part or a particular piece of the person that you need to put together. And we saw a great growth over a period of time, you know, in the last 15 years or so. But now the industry is changing.

Right. Now what we are seeing is actually the other way around where clients or operators don't really want to like you know

like hundreds of options to to bring a product to the market. Instead, they want a platform or an embedded offering that can just do what they need to do without the minimal overheads, without the minimal complexities, and you know, of course within the financial services, all the regulatory burden that it could have, right? And that's the trend that we are seeing. And again,

With the rise of AI, you know, there's always a question now about biosis: should you really buy an off-the-selled platform or should you build one internally? Right? So there are quite a few questions that come to your mind. And today we're gonna just generally dwell into these things, like you know, what happens when the cost of building goes down dramatically? what happens when generally the size of businesses are actually like in in hectum

employees count you know is actually reducing you know your revenues then the other metrics might be growing but the head counts are are are

Given that the industry generally is just very fast moving, you know, we are not seeing this kind of noise and happening in the last whatever fifteen hundred years, right? So the industry is moving really fast. Something you think is a cool idea today, but by tomorrow or in a week's time is already outdated, right? So how do you really deal with all that un uncertainty, right? How do you really launch new products in the market? particularly in a in a regulatory space where there are a lot of

Non-obvious things that you need to consider, right? So that's what we're going to talk about. But before I do that, let me introduce the panel today. Thanks a lot to both Philip and Alistair, who kindly agreed to be on the panel, and particularly Philip, who had to travel all the way from Berlin today.

to be on this panel. and I'm really glad you and thanks a lot to both of you. and let let's get started. But let's first do the intros what you do first. I'm Philip at Azure we help clients, typically non-banks, access the world of financial products.

at scale to drive customer value. So we typically move the most important KPI for business into the right direction. And we do that by partnering with an ecosystem about 50, 60 licensed companies, cards, accounts, loans, insurance, etc. And we bring this in an embedded fashion to the customer. And Alistair here is not just a customer but one of our favorite partners is how we know each other.

thanks for it. hi everyone, Alastair from Integrative Finance. We provide infrastructure also for businesses to build financial services, but if you can imagine.

Everything that we do is kind of the undifferentiated heavy lifting from the ledger downwards and plugging people into the ecosystem of regulated parts. Great speaking. Thanks a lot you guys for for being here today. as we get started, you know, let's we start with you, Philip, first, over your career, we have spoken in the past as well, we've done our life.

Few weeks back. and I I know that you have been talking about banking and embedded platforms for many years, right? What are the trends that you're seeing currently in the industry within your target focus or or outside? That kind of makes you feel that you know what the the vision and the reason you started Azure. think, okay, you know what? I'm glad I did that.

man. twice before becoming a founder. when when when when Martin Richard and I sat together at the very beginning, s we saw first of all the opportunity to build a layer above banks and then offer customers a v variety of different products, they be the right insurance product at the right time or

you know, an investment product. So we very early saw that there'll be tons of different suppliers, all point solutions with APIs, and that this will be available and need to get to a customer somehow. And th at that point in time it wasn't clear who that bridge would be. Today what we see is that Revolute has you know down to the ESEM card, right? And we we can also talk about travel companies doing the same, super apps doing the same.

people companies like money supermarket.com or check twenty-four building a bank to start bundling all of these things. So what we are seeing is the combination of a lot of these smaller products coming back together in a better way than they were before. So that's a trend that we see. It is a natural trend. Every product

Has this internal drive to be closer to where it's needed? I could give many examples so I don't speak too long, but what we're basically seeing is that the product is showing up where you'd expect it to be. And that trend is definitely becoming much, much stronger. Alistair. Now, a lot of people may not know, but I used to be a developer, like 20 years back, 25 years back.

I

I still remember writing like integrating other libraries and everything, you know, writing RPCs and SOAP protocols, right? Which most of these people wouldn't know nowadays. Yeah. and then the industry evolved, right? You built on API first foundations, you know, with the REST APIs, you know, did an amazing job in integrating multiple platforms into a consistent user journey. But again, you are seeing a big transformation within that, know, with the growth of

You know, MCB and and everything else, right? So in a nutshell, what do you go where are we and where where are we heading? well I think if I totally agree with what Philip just said in terms of rebunding the products together, where we play, there are kind of things, there there are there's a huge trend really where everything is continuing to diverge. And what I mean by that is regulations between bus like countries are diverging.

the the speed at which people expect payments is kind of getting higher, way more towards instant payments. And the the the API point that you you picked up

Basically, that didn't bring us a standardization layer, really. It promised to, but if you imagine, if you're trying to rebundle them all back together, every API API is basically bespoke. And even between services that are nominally similar.

The API is expressing those services in very different ways. And so it becomes incredibly challenging, one to piece something that's not really standardized back together, but also as these kind of big forces of divergency and especially regulation are kind of playing at the same time. They're kind of pushing what everyone needs to do and exposed by API to kind of a wider audience, I think. So it's at the same time as people want to bundle back together, so it's becoming harder to do so.

If if I may pick up on that, just imagine, you know, I always compared to the car industry.

And when it made the first Mercedes-Benz like the very prototype, he made every part of that car, That was like the bank that coded its own core banking in the 70s, And then everybody started picking up its own pieces, and you know, we got this what you were describing. And we are somewhere, I think, in the 30s or the 40s, where different car manufacturers buy different engines and different tires.

And the you know the accelerator brake clutch thing has not been agreed yet, and you can't take this clutch from here to there to this and that, right? But if you look at the trend forward, this convergence or normalization of services, PSD two or open banking, PSD three, I mean we all want it, right? So I think a general trend towards this is

is coming but it'll never be that you can just take the Volkswagen engine and stick it I don't know in a Ford. Yeah. I'm I'm I'm just thinking out loud here, right? with SAS it was all about

Like you know, breaking it down. And very rightly you said, you know, different parts are done by different companies and different platforms. But now with AI, you know, it's like, okay, you know what, why should I buy this? Like I'll just build it internally. You know, why should have another line item on my PL, right? and that also makes sense that okay, you know what, as an individual platform operator, it's difficult for me to sustain like a niece offering.

unless I'm like super specialized on it, you know. but for most sites are generic, right? and it's better for us to have an end-to-end solution that the customer can customize and deploy in their own workflows and all, right? So personally that's what I think is happening. But do you agree this is like a really a trend in the industry or is just like a like a hype? Well I'd say what I see is if I guess we would

You're talking about kind of horizontal SaaS where you know kind of generalized offering can serve many businesses. I think the winners of the last few years have been kind of very vertically specific SaaS, i.e., could be rainforesting kind of restaurant payments or whatever. Super specific. And they can build a much more detailed data model on those sorts of businesses, which is the most important thing as we move into this AI area, basically, like having as data as you can possibly have about your target customers is

Is the dream. But that's also allowing them to more easily bolt on extra things to resell to those same customers. effectively if you're a restaurant and you're using their software, you need to get paid. so that financial services element is a natural kind of bolt on it. They're able to bolt these things on more efficiently, I would say, than kind of horizontal and generic SaaS have been able to do. Philip, I have a question for you.

Asume, normally you talk about embedded finance orchestration rather than building the infrastructure, right? What part of that orchestration that normally your target customers or operator is underestimate massively? Whoa. well I I you know, I spent an hour or two with Lovable and I think it's all done,

So, first of all, the different parts aren't really the same. What one banking as a service provider does, the other one does differently. What one thinks is an account is not an account. A payment is not a payment. all of this divergent. And then what customers see is this, especially if it's done well, and Apple proves this all the time, right? The more well done, the better it's done, the less you notice it. The more

ubiquitous, the more transparent, the more embedded, the more natural, the more right? So if somebody comes and says, well, it's just an app, I can go to, I don't know, pick a country, Bulgaria, I can go there and get an app developer, 20 grand later I have my next revolute, right? They totally, totally underestimated what there is underneath I

Now call it kind of an iceberg. You see that little thing up there, and then there's all of this stuff down here. And to be quite honest, I don't know why it does have to be that big. Like at the end of the day, money, money there. What the hell? Like why is it like don't ask me how we got there? But we are there. Right? And but all of that stuff is important. Because it has to be secure. We have, you know, people who make law laws, Dora.

C I DSS, privacy laws, and they all make sense. And then you know, we are supposed to supply something that's legal, that is beautiful, and and and and and all of the rest of it, and that's what people heavily underestimate.

And Alistair, just building on the on on top of that, right? Now which a specialized like you know platform offering a particular solution, right, and customer integrating all of that.

of course that brings its own challenges, complexities, making sure what the platforms or tons of the platforms work well together. do you think? customers really moving away from it? with AI, you know, code writing, it should be easier to integrate everything. Or are they moving away and they just want a point solution that can do well good question. I think well, there are certain elements, right? You can't buy code as as as Philip said.

A ledger, a store of money, probably shouldn't be one that you VI code. We did have a potential customer who did it and lost money, so not a very good idea. key going forward really is systems of record in general, I think are going to be increasingly important, and we're going one level up from systems of record. What I mean by that is can you provide the

Well in the old world, you would provide a human user ways to access that system of record and interact with it. It's increasingly going to be agents. And so can you provide on top of your system of record a way for agents, both yours and externally, to be able to safely interact with that system of record? Because without it, they can't do their job. You know, business workflows are complicated. and it's actually to your point, it's usually multiple systems that need to be

to be connected together and so it's almost like thinking about here's my core product but I I need to be it needs to be able to be connected to all of the other things that makes the business run and allows agents to run workflows into the food.

You know what? kind of triggered me an idea in my head, you know, which with William might know. I'm currently going through and house, you know, it's in project and know, there's a lot of things that you you need to figure out, windows and doors and your flooring and blah blah it's a mind fit. And I met a builder who who said like, you know what, you can probably get a lot of things done on your own, right? But it's a headache and risk for you.

But just tell me what you want and I'll be the single point of contact for you and make sure everything is done. And if anything goes wrong, you come and hold my neck. Is that what's happening within this whole rebunding? That you don't really want to orchestrate a lot of different platforms. Probably you don't even know where the risk lies, where the compliance, you know, you're not meeting the compliance requirements.

sort of you trying to figure it all out. You are saying, okay, you know what, I'm gonna have a platform that gives you bulk of it and just let them make them liable for it. Yeah, absolutely orchestration is the key in this environment. Being able to coordinate across multiple platforms is absolutely what people want. And we and we see it specifically in the banking stack where point solutions at every level were incredibly successful as you described over last 10 years. But they're all iterating towards the same product like but either if you're a card issuer they're often back voting on

If they're a bank and they write cards, they're all kind of iterating towards the same same product so that they can control the end So what Alistair just said is is really important.

You're gonna have these agents that are going to do things for themselves, for you, for a business, right? And then you have artificial intelligence, and people still think in this kind of because we all used to use the web one app, right? Or whatever it is, and we start we still think in this paradigm.

Of there is an app and then there's a website and it does things, and then there is the prompt, right? And the prompt is inside of a web browser. But what it really is is a new operating system. It's a different human computer operating system. And this whole idea of somebody who's building a platform and operating a platform that will merge. So the the whole in five, seven years time.

We know orchestration will become more automatic. we will understand the topology of how data moves and we will interact very differently with these systems than we do today. the question is, I'm not quite sure how we get from there to there, I'm pretty sure about there, and I'm pretty sure where we are now, but I think that we as

in when we talk about this, we always talk about it as if there is the platforms, builders of the platforms, and the users of the platform. This is all going to get mixed together. yeah, I certainly So you know what, let let's take a a step back. You know, I think we're jumping too much on on doing doing the operational sort of thing. But let's say a customer comes to you, you know, could be a big e-commerce operator.

you know, in the UK, let's say next or governments for that matter. somebody in that boardroom says, okay, you know what, let's improve our customer loyalty or, you know, the R pooper for our customers, right? And let's launch this new offering, you know, emerald banking or you know buy now or whatever it is,

Now building it and integrating is one part of the challenge, which is becoming easy as we say with D AI. But apart from that, you know, what are other things that they need to consider? Like you know, operations, commercials, UX, you know, headcounts, you know, a lot of other things, right?

Philip, you know, what what do you think? You know, where is the Azure's offering here? And what do you tell your customers really? That's a good question. I got asked this today. the so what question. so it's great that we all know this, so who cares? I think

To be quite fair, really knows how what five years or three years from now feels like or looks like. I think if in Devon Hams they're going to say we want more customer loyalty, interpretation of that, what does it mean for the store? Are we gonna sell stuff more online? Is it going to be like this? It's going to be like this. I think people have to take quite big bets in the future. What we said at the beginning, you know that it's changing faster and faster. I think that large companies don't have core the core competency

Banking, finance, at all. They know retail or they know travel. So I think there is obviously a space for hand holding and learning from each other or providing services to that. But apart from that, what really driving customer loyalty will mean in five years' time, I think will be very different than what it meant today.

So I I think the con the consultant or the handholding will happen, but I I do not know exactly how the interface will be between customer, service provider, and the infrastructure providers below. Will it be a customer loyalty or an agent loyalty?

Anyway, so you know this you know agent personal agents are coming left, right, and center. You know, agents are making decisions on behalf of the users, particularly around how the products are discovered.

for the user, right? And agents are also like acting on behalf of the owners or the users in getting things done. And there's a big momentum right now around agent e-commerce and agent payments, right? because I'm sorry Philip, I wanna come back to you. With all these things happening, right?

How do think the like there are a lot of debate happening in the industry around product discovery? But in your mind, do think the users will discover new products or new offerings?

I think it'll be just much for for the us humans it'll be much more human. Right? It's already the prompt I can write it, but this will be voice, necessarily. But I always I'm trying to at the moment make a clip about embedded finance in five years time. So I'm thinking about, you know, somebody sitting in a car saying to the car, listen, did did you get my flights booked for the weekend in Vienna next week? Do we have a hotel booked?

Did you check on this and did you check on that? And somehow the platforms are just gonna take care of it and arrange for the payment. It sounds rather futuristic, but I think that's the most human way that we want to interact with technology. And I think it's gonna go somewhere into that direction. What do you think?

Well, I think again, talking specifically about banking and embedded banking, I think what will happen is Philip kind of described this already, like financial services are a separate thing still, and we're gonna increasingly see the merging of financial services into everything else. So it's completely behind the scenes. And the natural kind of if you roll this forward, effectively like financial services will be provided at the point which or platform

In which people spend their time. So Facebook, social media, effectively, people are doom scrolling and want to buy stuff, you have to embed financial services there to make that work, like seamlessly. And so you could see providers that would traditionally be customer-facing.

be shoved down the pecking order and be you know almost warehouses up to those kind of systems of interaction a bit like what happened with search where you know if you wanted to get yourself found pre-google there was a certain method of doing it. I don't I'm not sure I remember exactly what that was. post post and stuff like that. So suddenly one firm completely dominating distribution, right? That's potentially you know the the route here where

Assistance interaction like control distribution of financial services

No, I'm intrigued though, because we we seem to be revolvering around that the bundling point is what we used you said further to So that once the interface is really human, right, and the pe whether it's a social media platform, whether you're talking to your car, right, the agents are gonna choose the services. And at the very at the layer that Alistair works, we're still gonna have to make sure that the payments are instant, that they're regulated, that they're safe, that they were they're properly

actioned, right? And this is really a a really difficult part like

When is something sanctioned to actually happen, right? But I think the the the rebundling will kind of the agents will take care of it. rebund the agents will choose this service for that, this service for that, this service for that. As long as we can get the actioning right. And at the moment we haven't solved that piece. That we can say to an agent, go, yes, you can go buy something. go buy something, insure it, and then tell the trust layer. Yeah, the trust layer isn't there.

So is it is it gonna be like I choose this method of payment and I choose this platform for flights and I choose or is that gonna be some form of of advertising? Yeah it's it's gonna be preferential, isn't it? It's just that Google says I wanna advertise. So I have a person on that,

You're talking about. Yeah. Right. So let's say agents are making the decision. The user wants a particular outcome. Right. So let's say I I'm sending money from here to there, you know, this country to that country. is a is a given size. The agents figures out what is the best way to settle these context. And I don't care whether it's going through HSVC or through Barclays or it's going through

You know, this KYC or that KYC, you know, it could be anything, right? As long as the transaction is completed as per the whatever regulations are there, I don't really care where where is basically executed on. So what does it Sorry, can I just do do one point? Sorry, I was w I agree with what you said. However, like

It's the devil's like the the the big thing about payments and moving money is it's a mountain of edge cases, right? So in your case, what if I said or you said actually the sweater that I bought, the agent has bought for me is blue. In old world, I do a chargeback or a refund. But like how does that work when you can't prove that the agent didn't follow your instructions? Do you see what I mean? There's all of the stuff, the reversals and everything.

That would need to be taken in account of. So I wanna ask the question from the the operator point of view, the platform operators, right? So for them, now, like you know, historically or until now, if you are an operator, let's say a settlement network, right? Then you pre-integrate with you know the the UX layer or the acquiring side, and then you know all the transactions are going through you. But now

The agents can make that decision on your fly. So, how would you, as a platform operator, be relevant?

so first of all his point and so that means trust is still important and it's already happening. I I can't name names but I know some very big platforms in the UK that provide services to homeowners where they're being disintermediated because you no longer need to go to the site to get the referral that you otherwise would have gotten,

And in that sense for some things this works, but for edge cases when you said I don't care how the money gets there, I'm not so sure that's true because maybe with this method of money I can't get it back, I can't get it just the

There is a reason why things like Swift have existed for such a long time with such an old protocol, because people know it and trust it, right? So there's something that the H the how is the agents there's there's something more to it. and maybe that'll go away, but brand, I think, and trust is something as long as it's still vested in us, will still matter. And I I can have it slightly well.

semi-different take. I think for a platform business where you're sharing your infrastructure across multiple customers, in the in the agent world, I'll say this as a semi-controversial thing, but the really the only defensible business model in this market is accumulating knowledge faster than someone else. And if you're able to sit infrastructure and share that across multiple customers, you at least if you do it right are able to accumulate

edge cases as I call them and different things faster than someone doing a point solution by themselves. And it's kind of trying to set yourself up so that you can take advantage of all of the different use cases that you see on your kind of platform products, whatever, and make sure that you are learning and making your platform better and then hopefully keeping that inside and not exposed to the agents to be able to get themselves.

I I just wanna flip the question again on the other side, right? On the user side, right? And it is coming to to the banking part, right? So right now, let's say you are looking for a banking provider, you know, you look at all

same but then you pick the one that is closest to you or you know offer you one particular product that you really like or meets your need you open your account there and then you're locked there for most of your life right you most of them are not really going to switch right your money your wealth your income your expenditure everything is within the bank

Or within the agentic one, can this whole relationship can be flipped over? Where is me, you know, I'm worth something. I can produce certain income, you know, I can I decide where I wanna spend. Now you as a bank need to compete for my business. But they do. No, but right the big if we are still worth something. Yeah. So right now,

True. What are you producing? But right now, let's say your your net worth or wealth, whatever the measure is, right, is with the bank. So I can't access my money where I when I want. You know, I still have to go to the bank and say, hey, you know what? I need you to move my money and they will do whatever they need to do and maybe withhold that transaction. Right? But

The world I'm thinking of is the other way around, right? Okay. Like the old like in in many sense, like the old school way. But I'm carrying a my pocket full of gold coins and now it's with me, you know, and if you want this, then you got to do something for me.

You know for yeah. So I think it's very fascinating conversation. And way I see it that trust ultimately moves to the verifiable credential layer, machine readable, right? So then what happens is you've got these AI agents and every brand today has built modes, like you said, lock you in, all on based on human biases.

Right. So it's my favorite partly sponsors, the Premier League and every you know, like it's so I have this brand loyalty. But the agent has none. And so the agent's gonna look for the best deal and the actor might be out. Or not. The agent might be belonging to a company that have preferences for one deal because you get one percent here, whereas you will be getting point seventy five percent there and then ultimately you will buy a blue jumper with Apple Pay, with Visa, because the business Gemini

whatever it is, decided that this is the most profitable. But at the end of the day you still get your blue jumper delivered tomorrow. doesn't it doesn't matter the person who controls the agent has much more power than anyone ever before, which is why more than the banks, more than the customers, more than populations of which is why they all want the models. Exactly. But to go back to the thesis of our the great unbundling, isn't that the great miss said it.

the great rebundling, isn't that the great unbundling? Because every individual is driving through whatever financial services in different ways with different interests and so on. So if only at that demand hum

On behalf of human level, aren't we getting a complete level of fragmentation we have to deal with? The grey rebundling doesn't doesn't come to the to the actual user, it comes to the corporates that are trying to put all the services under one room. run from a financial service infrastructure provision perspective, got a completely different problem where fragmentation, on the demand side, not a bundling. well, yeah, you you have got fragmentation, but to serve the systems of interaction that I said.

Before you have to rebundle, you have to have a full stack product now. Where five years ago you could do a point solution and say it to many people, You've now got to have a full stack product because you're competing against global platforms effectively, you're not selling to a UK specific card issuer or something like that anymore.

If an agent happens to drop in and that's got a very specific niche, you know, payment need, that what is that what we'll end up with? Very, very niche stacks of solution. I don't know. I'm thinking making this up, you know, some some tiny, some very high value purchase of a god knows what, a a cross-border, sorry, personal jet with a Very, very specific that's this is what you is getting I don't know how it's gonna work. I have two mental models,

Agents are trawling websites and purchasing directly from a website, or it's coming straight through the the agents themselves, right? Like are they the new Google? So you give up distribution directly to your client, your corp, your corporate client that you're running owners for. Exactly. People say, like book the flights to San Francisco or something, it's entirely Chat GPT or Claw that does it. the grey the grey rib button can happen at multiple levels, it

So so let's let's step go back into that bundling and rebundling and all, right? Within the banking particularly, right? So bank old school, you know, so very like, you know, one set fits all model, eventually, you know, the rise of fintech and the SaaS and everything comes through. How do FinTech specialize into something? And then the customer can cherry pick what they want and build a product or a service that suits that particular business strategy and so forth.

We are now saying that things will go back into rebunded mode where you know they were looking for a point solution that can do it all. But the other side of the AI promise is to have highly personalized services. Yes. Right? So we see there's a call, there's a there's a contradictory movement happening right now. Why not both? Yeah, exactly. Why not both? So in your in your world, right?

What

do you think a new operator in this space? Imagine you're starting a new bank today, or a new, you know, a new marketplace today, whatever, right? And you wanna offer a fintech offering, you know, banking or a binary, or again, you know, or or a checkout experience, whatever, right? You know, what are the key learnings that they can take from the past, near past or long past?

well I think I there's only so many things you can do in financial services, right? There's five. pay and get paid, lend and borrow, save and invest, insurance, maybe rewards on the back of it, but that's mostly payments. Anyway, they're never gonna change, right? They're static. so that's a good thing, I think. What you're gonna struggle to do, I think, is as a new entrant, distribute.

those products to a very wide market like a traditional bank would say I'm going to create this one thing and sell it to 80 million people or something like this revolution. I think that's kind of the end of that. So you almost kind of pick a niche and instead of 80 million you've got to make it work for 800,000. Which you know where old school you have a bank where you have this huge fixed cost kind of technology and general regulatory kind of

cost base, somehow you gotta work out

How to and you can you can sell that to over eighty eighty million people so it actually works out okay. You've got eight hundred thousand people you're able to sell to, you've got to work out how to keep your costs very, very low. so the opportunity is basically financial services with no humans in it, or very few humans, where you know, old old world, you have a team of 300 in compliance that make everything working. You know, agents supporting a team of five humans is perfectly possible in the future, right, with the right

Of governance infrastructure in place to be able to enable that. So I think we've probably enabled a hundred between us, probably a hundred companies, right? And obviously not all of them made it. and I would say what I what I say now, or what I would say, is that it's not necessarily just about technology. Distribution is your number one thing. Do you understand your customer? Do you know who you're selling to? Do you have some kind of edge? Right? Revoluta.

some kind of edge. N twenty six had some kind of edge. N twenty six you're in Germany. Maybe some people only speak English. That you know, I mean like wow.

But there was an edge good enough to bring the cost of acquisition down to five to ten euros, which was you know a twentieth of what the regular banks were, bam, a new company could exist. So it's not it's not about if you're something new and you're gonna be somehow customer basic, it's about distribution. The technology is there, don't build it yourself, please don't build it yourself. Go to Alistair, come to us, we'll sort you out. Trust me. It is much more complicated than you think.

Whenever somebody comes out, I'm gonna build the next revolute, I'm like, okay, good luck. but Martin, my co-founder, has made a very, very good point. big, big banks, For them it's a different thing. is still going to exist, but what does lending depend on? Lending has, you know, how long you're in the market, how big is it, how much you know, how much equity do you have to play with.

how much can you influence government, how you know like you you can play basically the landing game, how good are your processes, how can you drive down the cost? And for those th for those players, size and scale and optimization, every little thing makes you bigger is better, right? And so kind of like distributing water, these as utilities will continue to exist and will not go away. But if you start from scratch, don't go into something where bigger

Better is better because good luck. you know, I want to pick a another good example on what you say. you know, distribution is the key nowadays, right? And we are seeing these with Next particular Next.com, you know, the online retailer. used to sell their own products in their store online and everywhere else. And lately, I think over the last three, four years, they started selling other brand products as well and becoming their distributor.

Right. So GAF, for example, exited the UK market and now sell through next online as well as in store,

Now imagine you are, you know, CIO or you know, whatever, some chief decision maker at next. And now you are saying, okay, you know what, we have this distribution power. got all these users buying from us, our product, our competitors product from us. you know, we are kind of a pseudo Amazon for the UK market, right? You know what, I wanna monetize more of my user, right? Yes, banking is what I wanna do on our next branded account, cards.

You know, next branded wallets that they can check out wherever and so forth, right? are the top one or two tips that your board will give to you that decision maker? I think the a live example of this in play is that Uber now issues Uber branded bank accounts to all of its drivers, proviso being when you close the door of the Uber,

You will get paid rather than you know to your external barclays account in five days or something like that and get instant payments. So the driver, that's incredible, right? And all that Uber wants to do is retain good drivers. So it's the it's the USP. However, it also enables them to earn an extra three, four, five percent on top of the platform twenty-five percent, right? So your minimum minimum, exactly. And then like times by however much they're processing on a daily basis, ten billion or something like that, is material to those

businesses.

however, as Philip highlighted, if you just look at the numbers of adding financial services, you think cha-ching without realizing all of the crap that's about to come your way in terms of everything going wrong all the time and people trying to scam you and all that sort of stuff. So it's it's balancing the decent economics if you if you can acquire users very cheaply against whether you think you can control the risk. It's one of the problems we have in the industry.

My co-founder is like, sales, sales, sales, sales, sales, sales. I'm like, yes, yes, yes. Well I pick up the phone and say, do you want some embedded finance? Like think about, hey, you want some embedded finance? Like what what are you talking about? What is it? Right?

And the thing is for Uber, like we can explain it, right? for big retailer, maybe it's not to be disintermediated by by the agents and actually marketing to them and whatever that means. maybe controlling the platforms and working with Google so that the agent's agent is being controlled, right? And maybe yes, it is loyalty and and cash back and all these kind of things you can put on a card, you can associate with something else. Maybe you'll see like crossovers with TikTok, right?

You want to buy a ticket? There it is, or maybe working with influencers in new and different ways. Again, the they promote your your trousers, influencer gets paid, gets you know, all of that is all embedded finance, right? But the difficulty is when it comes to what does embedded finance do, is well.

What can you do with a toolbox? Depends on what you wanna fix and improve. It's it's it's a Lego. It's a Lego thing. So you know if you if you think about it, every single business is kind of in the middle of the flow of the fund from one person to one to the consumer. Yeah. Right. and if we fast forward a few years, you know, given the pace of all this AI stuff happening,

What do you think, if you're a mid-to-large-size business where you do have a significant say or control on this flow of funds, what do you think they should absolutely be thinking of in maintaining their position or improving their position in the market? Don't ignore this trend. I was a tuber boss in 1999.

And CEO of this, I was working at Razorfish. we were building website, we were reimagining business. and we went and we said should you should do online you should have a store online. the posting is like, Well, how much is that gonna cost? Razorfish at that time, the website was minimum ten million, know, the servers, the T one line, the team, then whatever. eighteen months. And he's like, nobody's gonna buy close online.

Ten million, I'm never gonna I'm never gonna do this this trend. And I'm not saying Hugo Boss isn't around anymore today, obviously still an important brand. But obviously they have a website today, right? And maybe it would be bigger if they'd done it earlier. So what I'm say the question you ask is a bit hard to to s to to answer because it's a different answer for every kind of business. What it is, is it the speed, is it is it is it the agenda

That's coming towards you? Is it about loyalty? it about size of basket? Is it about optimizing your your supply chain? But all of those things have, like you say, flows in them. And I would again suggest come talk to us. And maybe we have an idea. And if we don't have an idea, we'll send you to somebody who I think if you're a brand,

Understanding your users is the most important thing you can do, and getting access to what they're spending their money on would be quite useful,

especially I mean we we we pitched to get a good gambling company and they weren't really interested in the upsides of embedded banking. They were only interested in the data that they would able to be able to see for their high rollers where they were gambling elsewhere, what they were spending their winnings on. They were like, this is not possible, is it? It wasn't possible. but I think access it's it's so rich and would like I mean if you track people across websites what you don't see is the purchase at the end of it if you've got the transactional data.

fits together. So it's it's additional data.

And actually, just gonna say every company is gonna be now fintech, including the AI frontier labs. Yes, because of what you said. Exactly. But but you know, there was a podcast I listened to, you know, just about this all the frontier lab and everything becoming like you know fintech company. But there's a parallel argument that ultimately AI is going to be embedded everywhere, you know, pretty much everything that we consume online and possibly offline as well, And ultimately the the value of that product or a service.

is is driven by how much computing cost you have in delivering that. So ultimately, if you do the projection further, then ultimately we are all just buying and selling and computing power and the cost of data. And and it cost of energy affects the Yeah, cost of energy. Yeah, yeah, yeah. Yeah. Yeah. So, you know, then then that changes the the the sense of the money, the value of the money that we have.

That is n possibly not gold, the cost of energy. We'll still go to the restaurant and have some pasta. Most of though. Yeah. Yeah. who? I was wondering you mentioned competition mean could we see like an arms race? heard Delta Airlines is deployed a new tech

To do bespoke pricing. I was wondering are all these agents going to neutralize each other

well prices. That's an interesting question because I think that is was there's someone once said British Airways is basically an unprofitable airline with a very profitable

credit card attached to it. system. Yeah. Yeah, yeah. And so yeah, they make a billion quid a a month or something from Airmars. It's ridiculous. So I think anyone who's got loyal customers and low margins, driven lower by potentially agents, you know, competing against each other, will need to look for ever other revenue sources to monetize those users, whether that be embedded banking or otherwise,

I think I think that's an interesting I don't know I don't know what the answer is basically, but like any sorts of businesses like that will will need to think fast. I mean we can we can pro potentially anticipate a few years ago it was all about Google versus Yahoo versus Bing. This is where we go this is where we're going. The agents are not gonna

destroy each other but they're just gonna compete each other and take customers from one side to the other because eventually you will jump into an Uber or into a Waymo and you will interact with another AI than your preferred choice and then all of a sudden you have an edge that this customer is actually a a a regular user Gemini and then he's moving onto an Uber or connecting Spotify or his Shazam or whatever and then all of a sudden there's a potential to acquire that customer from one AI to another AI.

There's always gonna be a little bit of an overlap at some point. You know whether you're in a British Airways flight, a new

Maulik Sailor (50:16)
But what I'm just trying to say, as I was driving there, I was thinking, man, most of these people are with TSB or they're with HSBC and that, so I think that, yes, at the very edge of these platforms, whether it's Amazon or Apple or Google or whatever else, for a great number of people, that we are always living at the edge. And that's why I made the comment earlier, we're still going to go eat pasta. It's the place across the street from where I live.

is a 70 year old Italian guy, does not accept credit cards, makes pasta at home, pasta is 12 euros, good pasta, and he doesn't care about AI at all, right? So I don't think it's an either or situation, it's just that choices are increasing. And the only thing that I think is true is that in general, the product always comes closer to where it's needed.

And that's regardless of whether you are in the old world, in the middle world, in the hybrid world or purely agentic. I agree for experiences, I totally agree. So holidays, booking a house, whatever. But it's tricky though. I mean, for toilet roll, I wish it just turned out for me. I disagree.

is learning what you have put in the past and why. Yeah, but fine, but then there's FinTech where the services may seem to most consumers completely undifferentiated, but the feeling of trust that they get from knowing under a number of operating with a familiar brand is something that's quite different. In fact, actually, I would like to put an important in between the customer and their work that they're doing can actually be counted for that too. That's an interesting point. think if Apple be able to help people trust Apple.

essentially

And

I think what happened here is we now have a new technical revolution that's just entering the financial industry. We don't need a solution for the people in the village who think RBS forever. It is a chance, yeah? So to see it as a mental program to take another 50 % of the market share of the existing ones. Or in other words, for innovative people to make more money.

So this is, think, the question here to discuss. What is the solution for people who are willing to invest? Cool. Let's have a look at the starting point. You mentioned M26, Monzo, and Revolve. Why did we go that way? Is it because we had multi-currency or currency exchange fees when we were on there? That was the reason for that, right? The other reason was, can I do a transaction in the middle of the night?

I can do that with HSBC now. Can I speak to somebody from HSBC at any time? I can as well, Homso as well. So the transaction, when somebody travels a lot or shops a lot internationally, the fees annually are huge. And that was a point of entry. But we didn't stop there. Now we have Google Pay, we have Apple Pay, and we potentially going to give you authorization to buy that BlueJumper that you really want. But that holiday thing, I think is maybe a bit published today. But AI is learning about what you've booked before.

And why they want your price range? So you will come with recommendation. But I think we're not there yet. We'll be there for a few jobs. Right. So before we go into more debate, I think what's always happening is it's always the same with the incumbent. could say, are these AI shift or this is the nowhere focus? So never work. Never ever. It's public bullshit, no trust, no nothing. Then suddenly 50 % of the market share are gone. And suddenly they all have been.

All right, I'm pretty sure we'll have a lot to debate right and you want the time I think It's about time for us to start wrapping to this discussion

But before we do that, before we do that, it's no longer called artificial intelligence anymore. It's called super intelligence as of this afternoon. No, apparently Trump made a speech at the UNGA and hands-forth is going to be called super intelligence. Trump? Yeah. Yeah. Yeah. It's not artificial. It's super, right?

Yeah. So anyways, building on that, what is the one trend or anti-trend that you are seeing that you're personally really excited about? Well, so I think there's a lot of hype, specifically about AI. Or anything. Yeah, well, so specifically with AI, I don't think it's really improving the lives of businesses quite yet. There's a lot of people running around.

But I don't think you'll see big productivity gains. You're definitely not seeing like massive employment yet. Vario and Odi said by now 50 % of entry level jobs would be gone. Employment's gone up. what I hope is that it, going back to the services element, if you can run a business at 15 million pounds of revenue where you used to have 75 people in the 7.

That should lead to a Cambrian explosion of entrepreneurship. So that maybe is the full case. I don't think we're seeing it show up specifically in business workflows quite yet in the way that everyone describes it. the hype would make you believe. Philippe, the trend I think is true is that it is ever easier to be the

to produce new ideas faster, to affect business with technology in better ways, and to move the needle. At the moment, I think we do not work with AI, or most companies do not work with AI in the way that they're meant to work with it. They're still stuck in, you know, windows and clicking and dragging stuff around.

They haven't realized it's an operating system. They haven't realized what an agent is. They haven't done any of that stuff, mostly yet. And we will see that. But it will be ever harder to build yourself a niche. Because the number of users is the same. It's just the number of ideas and the number of effects you can generate are ballooning. But that means that the number of users stays a little bit more or less the same. So I think it's easier to start, harder to succeed.

start by trying to replicate yourself as an agent. Start building all the... I'm going to show you something in a minute. Yeah. Okay, cool. So on that note, I would like to end today's discussion and I'm sure we'll have a lot over drinks and pizza. So thanks a lot, Alistair. Thanks a lot, Philip, for joining us today. Thank you. It was fun.

Sorry, just a few housekeeping before we end Max the next slide. So, you know, we have a few interesting events lined up as part of the Get Featured in the Community. There is one like AI Native originates and then landing, particularly on commercial mortgages that's coming up in October. We've got interesting panel on that and very interesting topic we're going to cover.

We have a live podcast coming up next week. That's about regulation and licensing for any financial products next week. And separately, we also started doing like 5K city runs. So the idea is to run 5K around the old square mile covering all the 12 gates of London, which roughly tends to be about just about five miles. Five K, sorry, five K, not miles. And we did one last week and we'll be doing one again in October. So.

Please do get involved. I'll my agent. I'll get my agent to run for me. Cool. And right, so just coming back to you, there should be a play button next. On this? Yeah, on the image. There should be a play button.

Is it PDF or online link? PDF there won't be. You need to open the online link.

Right, so what you said, we are actually experimenting something about cloning yourself. So you can clone yourself, build your own private knowledge base, define your own workflows and guardrails, and get your clone to do work for you. And I've been trialing lately in doing some meetings and follow-ups for me. And so if you want to give it a try, feel free to go to xbito.com.

Yeah So far I have managed to get it to attend meetings and follow a chain of parts Well, in a way, yes, and I'm also changing some of my Calendly, you know meeting links to the clone meeting link so, know some of the initial, you know information gathering can be done

Let's check you're really here. Yeah, that is the exito thing. So we've been trying and of course we is a project of mine. So I would love to hear your feedback. How would you use that and whether it's useful or not? But I think it is. At least I'm finding it's slowly useful. OK, anyways,

and that's about it. Thank you very much for joining today and thank you for a wonderful panel. Thank you. Thanks a lot.