Alex: (02:11.566) Hello to all listeners. We’re here in London kicking off the new season of the Curiosity Code podcast. Today we’re getting into the real challenges behind growing fintechs, why so many digital transformations miss the mark, and what happens when messy data and GNI collide inside financial markets. With us is Stephane Malgay, board advisor to fintechs like TransFICC and Neptune Networks, and former senior executive at J.P. Morgan, Société Générale, and ING, where he led innovation, trading, market structure, and digital transformation. Stephane’s seen it all—how fintechs scale smart, how bad data can kill a strategy, and how GNI is already shaking up the future of finance. Stephane, welcome to The Curiosity Code. It’s great to have you here in London. Stephane: Thank you, Alex. Thank you for inviting me. I’m glad to be here with you today. Alex: Let’s dive in. I have so many interesting questions that I’d like to talk about today. And I’d like to start with your role as an investor and board member. Do you know any internal signals not seen on the pitch decks that tell you whether a startup is building a true company or is just a clever feature? Stephane (03:33.868): Yes, of course. Thank you for this question. I think what is important is the way we invest in a company from a banking point of view—because what we found is, you find a lot of great ideas in the market. Some people come with experience, they see a gap, and they want to solve that problem. A lot of the time, because they are early-stage startups, you have to help them grow. And when we realize that, we’ll say: You know, there are different ways you can help them grow. Either you become an investor—so you give them money so they can survive. Or you give them data, for example, so they can prove their hypothesis through the data the bank has. Or you become their client. And a lot of times, it was all three. Or we started by wanting to become the client, but because of that, and their stage, we gave them data. And because we liked them, we said, “Well, let’s invest in you because we helped you grow.” So I think it’s very important, from your question, because what I found is: having a good idea is not enough. And we made this mistake many times. I was part of what you call the innovation team on the bank side. You can find lots of people with great ideas. And that’s what the pitch deck is going to be about, a lot of the time. They explain the problem statement and how they want to solve this problem. And it’s fine, but it’s not enough. We had a lot of great ideas within the bank as well—things we wanted to change and improve. So if that’s too much, at least we tick the first box. The second thing that we looked at is more the team itself—the fintech team: how they’re organized, who are the founders, what is their background, do they know enough about the subject matter? Do they bring people within their team? Not everyone in the team needs to be an expert on the subject, but at least they should bring different skills. So you may have a team of five people. The CEO may come from market experience in a bank—he knows what is broken and what needs to be fixed. He finds a great CTO who helps him from the technology side, knowing what new tech we can bring at a lower cost as a solution. Stephane (05:56.438): And then they may have someone more on the marketing or business development side to help shape the company. So, the role of the team was critical for us to say, “Yes, you have a good idea—but do you have a great team?” That was the second check. The third check is really trying to say: even if you have a good team and a good idea, do you know how to execute? And I think this one is probably the hardest to have and to test. So many times, someone came in who was a Managing Director at a bank, had a big job, etc.—then goes to a fintech company, brings good people, but they never executed themselves. They were head of 600 people, and within those organizations, they had people executing for them. But when you start in startup mode and you only have five people, you cannot delegate everything—there’s nobody else. Alex: Exactly. Stephane: So you have to have more experience on how to make difficult choices. How do you execute a strategy? How do you fail and recover? How do you pivot? So that’s really the three things we looked at. Stephane (07:20.334): You also think of yourself in the early days—you know, the idea, the team, and the experience behind it on execution. If they have those three—though many times they only have two out of three—then we had to take a risk. But if they have all three, then for us it was a great sign: okay, these guys know what they’re doing. They know the problems they’re trying to solve, and they’ve done it before. They know how to execute. Alex: Yeah, I think the execution part is very insightful—at least for me, hopefully for the listeners as well. When I was listening to you, I noticed something: your first point was about experience. Usually, like you, someone comes from the banking industry—maybe as a manager running a department or something. Then I find that the execution skill and that kind of contradict each other. How do you kind of measure execution skill—or do you just trust your instincts? Stephane: Well, I think that’s where you can partner, right? The person who understands the business and the problem statement may not be a startup entrepreneur himself from the past. He’s moving from a bank with 20,000 employees to a company with five employees, so he won’t have that execution background. But he can find someone to work with him—like an ops person...who was already in the fintech world, created something, was successful, and wants to do the next big thing. That’s when it starts to work like magic—because everybody brings something very strong to the table, but from a different angle. They can work well together and execute on those strategies. Alex: Let’s talk about scaling—because you’ve helped scale lots of fintech startups. What’s the hidden danger you’ve seen happening across fast-scaling fintechs? What are the hidden traps teams fall into? Stephane: Yeah, and I think I can speak from my own experience. We suffered a lot and learned ourselves the same way—because we had companies that we were trying to create in-house and then scale up externally. We also worked with fintech companies from the outside that we were bringing into the bank to help them scale up within the bank. What we found is, probably like with many others, a lot of them failed. And when you're in innovation mindset, that's how you learn. Not all projects will succeed, but you have to learn how to pivot and stuff like that. So come to mind,Some of the mistakes I saw in people who failed to scale were: First, they had a good idea, they had a good product, they found good investors to support them—but they were burning too much cash too fast. On the pitch deck, they often say, “Within the next three years, this is where we’re going.” But every founder knows, it’s never going to happen that way. It’s probably going to take six to nine years to get there. So, if you burn cash based on a three-year trajectory, and you realize midway that you're not going to make it, you run out of money—and then you fail. That’s a very common mistake: burning too much cash too quickly. Then the investors say, “You didn’t prove the value of the product, so we’re not putting in more. So be careful with that. It’s important to have a long runway and to be able to measure your progress along the way. If progress isn’t happening—there may be good reasons for it in year one—you need to course-correct. That’s the first thing. Stephane (11:06.126): The second thing on scaling: a lot of times, you prove your value with your first client or your first two clients. But if you need to scale to 200 or 2,000 clients, that’s a very different story. Your marketing strategy, your go-to-market approach—everything changes. In your first two clients, you can babysit them. You take the time to onboard them. But if you’re onboarding 100 clients a month, you need strong processes. You have to automate that onboarding. You need scalable marketing strategies. In our market—just to give you background—it was on the investment bank side, not retail. So it’s more professional markets. Unlike building a Facebook-type company with massive user bases, in investment banking, even reaching 100 clients could take five years. So if you succeed in onboarding 100 clients, you’re doing very well. If you want to penetrate the banking system—or at least the investment banking world—it’s very hard. It may take 18 months just to get through procurement. So my tip for people scaling: plan in advance. It’s going to take time. If you want a bank to be your client in two years, you have to start speaking to them today—show them what you’re building. Stephane (13:28.622): And one last example on scaling: You start with a small team of five. Everyone knows each other, they work well together, they’re passionate. But as you grow—25, 50, 100 people—you bring in external hires. You don’t know them as well, and they might not be as passionate as the original team. So, it’s important to have a process for scaling company culture. Otherwise, you risk degrading the client experience. By the time you reach your 59th client, they might say, “Sorry, but when I call, no one picks up the phone.” And that’s when problems start to emerge. Alex: It’s very interesting, because most of the points you covered are applicable even to my business. I’m not a conventional startup, more of a service provider. But from sales cycles—which are also very long in financial services—to the challenges of scaling, it's very relevant. So I think back to the point about execution and how it’s a key factor in startup success. I think we can even broaden that—execution skills matter at the business management level too, not just for startups. If you’ve run a small or medium-sized company before, you probably know how to handle cash flow planning, how to scale processes, how to scale operations. So it definitely seems to be a critical factor. And it’s hard—I think a lot of people don’t realize, maybe that’s a good thing—when they start a fintech company, they don’t realize how hard the first two to three years are going to be. How much effort it takes. Compared to a 9-to-5 job, it’s a 24/7 job. And in year one, they realize, “My God, this is hard.” They have to run the course, and then they’ll get the benefits. But the first few years are pretty rough. And some people give up—they say, “It’s not for me anymore. I didn’t realize it would be this hard.” Only the passionate people succeed in that game. Alex: My next question is related to what we’ve covered, but from a slightly different angle. Suppose a team has strong expertise in their product, in business management, in the market where they want to launch. They come to you as an advisor while raising capital and seeking guidance. At what point should they stop listening—or at least listen less—to external advice and rely more on their internal instincts? Stephane (17:25.773): That’s a great question. I think it’s a mix of both. I’ve made that mistake myself. The role of a board member or advisor is very different from the role of a manager. A manager trains a team and can direct them. An advisor, on the other hand, helps the management team go to the next level. You can’t really tell them what to do. You have to work with them to help them think and grow. And they don’t have to take your advice—because it’s more about guiding their thought process. What I realized very quickly is that the CEO has the hardest job. You can probably relate to that. The CEO is very alone. Everyone in the company looks up to them and says, “You’re the captain of this ship.” So when problems arise, they all come to the CEO. It’s a lonely job. Stephane (18:46.350): And usually, it’s not your wife or girlfriend who can help solve those problems. That’s where the advisor comes in. A good advisor brings skills, listens, and says things like, “Why don’t we try this? Have you thought about that?” They’re not supposed to say, “You’re doing it wrong.” Instead, they help you think differently and open new doors. Often, in a board environment, you’ll have more than one advisor. One might have a great network of potential clients. Another might know all the relevant tech providers. Not everything needs to be built in-house—so you’ll want to know how to partner, how to plug in what you need. Each board member or advisor should serve a specific purpose: to bring their network, their experience, or their relationships to help you. They become your confidants. For example: “We’ve hit a roadblock with three clients in Australia. They want us to go there, but we don’t have the budget. How do we solve that?” And the advisor will help you figure out a solution. Alex: Is there such a thing as too many advisors? Stephane: Well, yes or no. I think you have your core team who is close to you, like the regular people that are part of your board or your advisor. And I think this you should keep it quite small, right? So people you can really open up to say this is not going bad. And then I think on the flip side, yeah, you can have a large network of people that you can go to for specific demand. Like experts. Right, experts, right. I want to find something on cyber security. I don't know, is there somebody who can help me, advise me on that? Knowing the market, what are the solutions we should look at, etc. So this you may want a range because you may come to them only once a year or things like that. But the close relationship, it's your board. It's your two, three, not more advisor there, I think. Alex: Yeah, I like the concept of seeking help from experts who’ve gone through the experience—so you move way faster than doing it alone. Stephane: Yes, and it’s impressive. If you take the right approach, many people are ready to help. Sure, the role of a fintech founder in the early stage is hard. You face a lot of closed doors—new clients, product development, everything. But you also meet people who want to help because they believe in your idea and want you to succeed. A lot of the time, they’ll do it for free. If it becomes regular—like you’re calling them every week—they might say, “Okay, now I need to get paid.” But initially, you’d be surprised how generous people can be. Just last week, we were trying to build a relationship with a large company. I messaged someone on LinkedIn: “We’re trying to solve this problem and need to talk to you.” He replied, “Let’s have a call tomorrow.” And it worked. I didn’t know him personally—just mutual connections. But we made it happen. So reaching out for help—it’s a game changer. Alex (22:35.916): Yep. Let’s move on. Stephane, across your journey advising and investing, what’s the most painful or expensive mistake that you can recall—and what would you have done differently if you had a chance to rewind time? Stephane: Okay, I’ll give you one. It was a long time ago. It wasn’t a startup—it was within a bank, a large project. My career started in technology. I was a project manager, developing new products. One big lesson I learned was the importance of alignment with the sponsor—the person who’s supposed to “buy” your product internally. We worked on a complex pricing model application for financial products. The project lasted two years. It was for the head of trading of that product. At the end of those two years—when the product was ready to launch—the sponsor left the bank. Someone else came in. And that person said, “Sorry, I don’t need it.” Stephane (23:22.088): That was one of my first jobs. I was proud to be part of it. We had a team of 10–15 people working for two years. And then boom—it was all gone. The new guy had different priorities, didn’t want to price the instrument that way. The finished product never saw the light of day. It was expensive. And painful. But not an uncommon story in big companies. Looking back, I should have avoided a single point of failure. Maybe I didn’t engage other people who could have supported the product direction if the sponsor left. It was a big lesson. Later in my career, especially when focusing more on innovation, I embraced a different mindset: don’t be afraid to fail fast. In that first project, I should’ve seen the signs. Now, in innovation, we celebrate when something fails quickly—because then you can pivot, move on, and reallocate resources. If you wait too long to fail, it becomes more painful and expensive. Okay, I will give you maybe some... I think I cannot tell. No, but like one mistake, this was long time ago, et cetera. So it was within the bank. So it was not a startup, but a large project. My career started in technology and I was a project manager and developing new product. Stephane (23:22.088) is the alignment with the sponsor, with the people who want to buy your product. We worked on a complex pricing model application for a financial product for two years. And it was for the head of trading on that product. at the end of those two years, when the product was ready to launch, And this was one of my first jobs. So I was very proud to work on that and to be able to do that. Um, but the guy left the bank and somebody else came in and, uh, it was expensive because it was a team of probably 10 to 15 people working on that for two years. And then you guys say, well, sorry, I don't need it. Uh, and the product was finished. because it's a tiny unit. Yes, because there's a new trader had a very different vision different ideas He don't want to price the instrument the same way and and so we came with a finished product because it was just by the time the other one left The product was finished. We are going to roll out to production and the new guy said well I don't want it and and for the team and for me it was like very painful because We gave everything we had for for for this product to make it ready to launch And this product never saw the light of the day. And in fact, in large companies, happened quite a lot. I'm not the only example there, but I think there must be better way to do that. In this case, I don't know what would be the better option. Maybe I didn't sell the new product well enough to that new guy. I didn't have a single point of failure until I had one. Stephane (25:14.956) clients and to, to, where I should have more than one, maybe some people who can help us into the direction, but it was painful, but it was probably one of the biggest lesson for me to not try to do that mistake again, as a product manager, and to make sure like when we develop something, you know, and this I learned, probably in the last five years when we were focusing more on innovation is you have, not to be afraid to fail fast. Maybe I should have saw some of the sign in my first project. So now in the innovation process that we had at the bank was trying to celebrate when you fail because that's the faster you fail, the better you can pick up and move on. Right, if you wait too long until you fail, it's going to be harder and harder, fail and more expensive mistake. And the analogy I use a lot is like, if you look at a baby trying to learn how to walk, you expect the baby to fail and he's picking up and he fell and pick up and up, but that's best way for him to learn how to walk. And I think in the fintech world it's the same thing. We should not be afraid. and so many people are afraid of that and they try to push. And that's why I use my example from early in my career. I see a lot of fintech who try to push their products hard and hard. And even if you start to see some signs that the product is not fit for the market they are trying to get now. So maybe it's too early, the timing is not correct. maybe it's too expensive, it's a very well-solution design, but too expensive for the problems they trying to solve. But they're still trying because they believe in the product. And it's normal, as a CEO, as a founder on tech company, people have to have the passion and to believe that their product is based on the market. But you have to listen to those signs to say, listen, maybe the market doesn't need it now. And this I saw something, another example, where Stephane (27:36.13) the same product, somebody created it like five years ago and had to switch off his company after three years. Somebody created now and they are the king of the world because the timing was correct. So you have to listen to the sign to be able to avoid the experiencing mistake. And I think with those innovation methodologies that we have now, you try to, you know, validate the problem you try to solve with different clients. to fight fast, to not develop the whole product day one, you do proof of concept, so pilots and framing just the visualization and try to see it with people. There's a lot of tools that we didn't have like 20 years ago to be able to test your market before it's become too expensive. I think another important aspect, coming back on your parenthood analogy with the kid, when you have your first child, everything is new to you. So you're getting really, sometimes too emotional, too unreasonable, too attached in a sense. You get the second one. OK, it's a whole different story. So my point is that. when you do your first startup, first product, and I had my startup like back 10 years ago, and I remember the feeling, it's like your baby, you're getting so attached, and you cannot accept the fact that you make a mistake, it was wrong, like what we call in project management, some cost. Okay, move on, close it, move on to the next one. Now you would hold to it and okay, I need to make it work. And I think that that's what you... you are making. Stephane (29:23.606) Exactly. And the banking world is extremely hard because like I did 30 years in banking and you know, if you fail, you get fired. That's that in the banking industry. Only the guy who doesn't fail get a nice big bonus. The guy who failed get fired. So so people coming from this industry to the tech industry, don't. If failure is not a problem, you don't fail. When facts. So failure is not an option. Stephane (29:52.664) Failure is, know, failing fast is the only way to progress and to make sure you are testing the right hypothesis. So, yeah, I think. I'd like to chat about digital transformation. You led digital transformation at ING. In your experience, what's the... real truth behind failure of digital transformations even when companies announced they as a success publicly. Yes, and I saw a horror story in my career on digital transformation because to some extent, like, I think this is important, because the electrification of financial markets has started with the internet boom in the year 2000, so 25 years ago. So still quite a long time to learn how to electrify our markets. But what's fascinating is the work is still not done. There's still a lot to do. is still pockets in our industry where people are still using fax machines or emailing paper on PDF and stuff like that, whereas they could do everything digitally. So there's still a lot of work to be done. But when I started in technology in my career, 70 % of the projects Stephane (31:25.238) were never delivered. The number was very high. And digital transformation is about that. How do you automate some processes within an organization, either building yourself in-house or through external help with fintech companies? over time, I hope I became better at doing it because we learned some tips of how to do it. The project management activity of such a program was critical to the success of what we did at the bank. So making sure you have strong sponsor alignments and something from my impact, think more than one sponsor, team of people who can review the strategy, agree the strategy and we want to get digital. And if one of them move, At least we have the other one. So alignment with sponsorship is critical. Your process to review progress is critical as well. Making sure you have a So we put in place a very strong governance model where we try to gather. So in our digital transformation program, we have a lot of different components and each component we are doing at its own pace. So, and reporting the, what we ask them to report is their progress, but what we try to push is to ask them to report what was not going well. Same thing, try to learn how to fail fast. It's like, tell us what is blocking you to progress. And this was kind of new for us at the bank. Because if you report that to your management, same thing, your management can become your advisor to say, okay, we are going to help you to unblock this issue. If you have a team waiting for another team that didn't deliver what they're supposed to do, or you are waiting to sign a contract with the vendor to be able to access, how do you unblock those issues and try to use the power of the management team to say, can you help us to unblock those issues? So this was extremely successful to do that. So not be afraid to escalate the issues that you have quickly enough. The other thing we did, which is... Stephane (33:46.658) you know, in banking industry, have two types of banks. You have the big bank who like to develop everything in house. They say we have large technology team and you have other banks who try to buy everything from the outside because they don't have. What we try to do is create a mix of both worlds to say what is critical for the banks that we should develop in house versus what is less critical. It's not an IP of the bank. So we should buy it and not be afraid to buy when you need to buy and not to be afraid to develop things where you think it's a differentiation. So pricing methodology, maybe something you want to keep close to you versus connectivity to different exchanges is something you could buy. So trying to make those. And then... not be rigid on that. So our digital transformation program was kind of a three year strategy. And this is like similar to fintech is things happen and things change. So you have to be use your governance to be able to change when you need to be and to be able to say like what we agreed three years ago is not true anymore. The market has changed, the market structure has changed. is new trading when you come into the markets and you exchange. is coming to the market, we have to connect to them. So be able to pivot even in a large digital transformation program, I think it's critical for the success of it. And then, yeah, it seems to work really, I think by doing those different tips and monitoring very closely the governance of that program, it managed to work out for us. We've been talking about how banking industry is really conservative in a sense and with long cycles with challenges in terms of execution and delivery. And it all, in my opinion, goes down to mindset that people have been having in this industry. How do you... Alex (36:04.096) approach that so you want to push the innovation you face this old mindset where data been thought about thought as cost center not like a driver of opportunities and changes how do you manage This was probably one of my hardest things in 30 years of because something like, if you think about it, how the bank work or used to work, and I work mainly on digital transformation, electrification of market for a long time. When I started, let's say in 2000, I was in another bank and we are trying to create the first electronic trading system for that bank. the challenge was that even on the management side, you have in investment banking, have sales and trading. was two big poor houses. You know, the sales talk to clients, try to get the deal done and the trader creates a pricing methodology and the risk management activity for those transactions. So in the year 2000, the cell would say, well, I know my clients. I know exactly what they want. And if they need something, they call me. I don't need an internet guide telling me to book something online. And the trader, something will say, well, I have my spreadsheet and I do my macro and my spreadsheet to price my instrument. I don't need the technology guide to automate my process. So changing the mic for me for the last 30 years was the most difficult thing to do because Stephane (37:43.758) You have those people who are very highly paid, who are very senior in the bank, who have a lot of experience, probably more than you, who have a relationship with top of the bank, and they don't like change. People don't like change in general, And you have to force them to say, listen, the world is changing around us. We need to be altered. So this was true 25 years ago. It's still true today. So in my last project, was the same thing. Some people would say, well, we don't want to change because... You know, we already have a process in place. seems to work. There's no reason to change. So to combat that, was a few things we tried to do. One was... that work well for us is trying to show them that the competition is already changing, right? You don't want to be the last one trying to board on the boat. So the fear of missing out, if other people are moving, how do you do that? So a lot of time, you know, we have very few people who want to be the first one, the change leader. It's very hard to get, if you get one of those guys, fantastic, he's going to support you and drive the change. But a lot of time you have changed followers who are going to wait until they look, they talk to their friends in other banks, they're using that, know, like trading digital asset in the bank, for example, is saying like, I don't want to be the first one, but if everybody else is doing it, probably I should do it. Right. So those kinds of questions. So trying to show, to get information, a lot of my work was trying to... gather information, what's going on outside, like to see the trend is starting to move and people are starting to, and data is probably a big thing. AI is another big thing is you don't want to miss out. So when do you start to do that? Second thing is to get support from top down, from the top management. So if you have a CEO who is very visionary, it helps because this guy is going not to be afraid to say, I'm going to check that boat because we have to prepare from Stephane (39:52.366) a digital way of doing things. Sorry guys, I you don't want to change, but I'm going to ask you to look at that and push it. So that was second key of success, trying to get somebody very high into the organization to believe in you. And if they do, then it helps a lot because you can use their name, you can use them to call out some of the middle layer which is blocking you to progress. And then it's... trying to prove the value as well. So that would be the first thing and you can do it smaller, right? So say if people don't believe in using the data because they think they know their clients very well, how can you prove that value? So we did as part of this, those innovation projects, we did a lot of pilots where we try to work on a very specific use case with a FinTech company and we gave them like three weeks to prove value or. One of them we did from ID generation to production in 16 weeks, where a lot of the time in banking to go to production in banking environment is going to take three to five years. In 16 weeks, we have something that could demonstrate to salespeople in production using data, some insight of their clients that they didn't know about. So that was very concrete. We did an assessment, say, well, you know. how much time this solution gave compared to... And the guy was shocked because somebody was that full in the beginning, but after those 16 weeks, was like, I want that, because it's going to help me to go faster, better, cheaper for my clients. So it can happen, but definitely it's hard. From all the digital transformations that you've witnessed, your opinion, what's the smartest, most underappreciated move that a company has made and that made all the difference? Stephane (42:02.422) So I think it's about being curious and not afraid to, I will use an example of one project we did, which was on data and AI, where we wanted to be something a bit disruptive on the way we communicate with clients through chat systems. And we didn't know yet if a solution exists, if it could be done. And this is hard because a lot of time, I said before, people are more trend followers and things, and we wanted to do that. The risk that we take and that we took in that space is that, and we had the support from the top, so we are happy to go for it and try to prove value there. And... To be able to prove that value, what we did is we found a very small fintech company. think there were like three people at the time working on this product. And we liked it and we said, okay, maybe we should try to work with them. But there was a perception that it was probably the biggest risk for a large bank to bet on a three man small tech company. So what we did in this case was we said, fine. We are going to ask, I don't know if I can say names, but one of the biggest tech companies in the world to look at this. One of the biggest financial data company in the world to see if they have a solution. and this startup just have a few people. And we gave the three of them the same input. said, this is the history of chat we have with our clients and want you to analyze this data and try to come up with what can we do with it. Stephane (44:17.154) And this was fascinating because the biggest IT companies that had compute power, like nobody else can have, they had like 500 data scientists. But they didn't have the expert knowledge of our product, or the financial product, right? They were a big tech company, but not a financial technology company. So if in the chats we talk about Apple, they will not know if it's a fruit. or if it's a stock price, right? So just using an example. So they were good technically, but not in the product sense. The second one was something very large company, but in the financial sector, so they could understand the product. But because they were very large, they were quite slow at adapting to change, right? So into a time compressed period to do a pilot program and to be able to... evolve through the data input that we were giving them, we saw that they were making progress, but not as fast as the small companies. So that was the power of the small fintech companies that they could, they understood the problem very well because it was an ex-trader coming from different banks. knew the problem. They understood the technology very well because they had data scientists within that small team who are very good and they could react very quickly to change. because they were nimble and they could adapt very fast to change. And so we learned a lot from that experiment because originally you want to go with the big guys because they say safe bets, know, there are big companies of 400 data scientists, should know what they're talking about. But at the end we want with a smaller company because, you know, reacting to change was very critical for us. We had to... take a bet on them to make sure they continue to exist in five years time. We didn't invest in them at the time. I don't remember why, but what we did is we're to help them to get other clients to say, you, if we are the only client for you and we are not, we are client number two, think. So, but it's not enough for you to succeed. You need at least five clients. So how can we use our network for you to find other clients and we help them. Stephane (46:42.062) to do that, to help them to scale and to do that. But we took a bet on them. And I think back to your question, this is what very rewarding experience for us and for them. It's because we did the test and we took a risk, but it was a calculated risk for a specific project. And we gave the chance to a very small startup to rise and work with a large bank and now several large banks. And now they are doing quite well. Sounds like a beautifully executed startup idea. A few people could make a significant change. And a lot of the time, for me, it's why in the bank, my role was to try to work with fintech companies and to bring them to the bank. We saw that again and again and again. It's like, if you try to develop in-house, yes, you have your own IP, et cetera, but the time to change is quite slow. Fintech company by design will be more nimble than a larger bank. and they can adapt to change very quickly. And in the environment where we start to see around data, AI, generative AI, you need to be able to react really fast. There is new technology, new tools coming out every single day. partnering with fintech company, even early stage fintech companies can be a game changer because it's going to increase the speed and you are going to learn a lot from them. We've spoken a bit throughout the conversation about data, and I want to slowly get into that subject. As many people say, data is the king, especially given Gen.Eye that's relying heavily on how good is your data. Alex (48:37.974) And unfortunately in many organizations data may be not in good shape. It's shocking. It's shocking from the inside. So, yeah, this battle that for... So given your experience, what you've seen in banks or fintechs, hopefully, you know, fintechs are good with this. I don't know. What is one thing that they could address, you know, first as first and why? So you can split into two sides. What is fascinating is on the banking side, finance banks have a lot of data, huge, huge amount of data, but all the transactions, they do the clients, the trading price of all the instruments, plenty of data. The problem is they don't... get the value from those data. Those data are stored in different systems. In one of the places I work for, we had eight data stores for very similar transactions. are storing eight different data stores. And we were not getting the value from those. So I think that the first thing is to try to, even before you want to clean your data, and this I agree, lot of the time the data is not Stephane (50:10.542) clean enough, you want to try to see what problem you can solve with the data you have before you go on by external data. Try to shape up the program, try to do some pilot, try to analyze some of the data that you have. Then the second stage will be okay now that I know I can do something with it. And maybe you don't have 100 % of the scope and you need to bring external data to that. And we'll get back to that. Then you have to clean it and to bring it back to the user. And then you can start to realize the value thing. And you can start small. It's just like everything I said so far is you can start small and then grow through it. A lot of just be more concrete to give you a simple example. A lot of banks will have transaction data for regulatory reporting, for risk management activity, for valorization of their portfolio, for compliance checks, for monitoring that your trader don't do anything that they are not supposed to do. But they will store all those same transactions five different times just to get a different outcome. So it costs five times more. And it's the same kind of data set. So if you can have a good vision first to say, okay, I want to this data to check if my trader are doing something they should not do to also do risk management, to also report this data to regulators. Then you can really, really reduce the costs, get some savings. And then you can reinvest that to create products out of it. And that's why, you know, I saw that many, many times. Now I moved to a tech company with data focus and we try to use public data for people to do that and face the same challenge but we create an opportunity from it is like people, lot of the times they give up too quickly. They will say, oh my God, we have like five data stores, the data is not clean, et cetera. Where I work now, we consume data from 65 different places. Stephane (52:34.318) We clean the data, we make it unified and we present it to a very simple way for people to digest. Right. And then it's the war factor, but we did a boring job for them to some extent. And that's a lot of people are afraid to do that because they say, Oh my God, it's like, you know, you arrive to a messy kitchen and you're like, have to clean everything. I do it tomorrow. Right. It's sometimes you have to take the step to say, if you want to get the benefit out of it, you have to do the boring job first. and then you will get the benefits out of it. Yes. So it sounds like it's not always about amount of data and the quality of data, but sometimes it could be that it's excessive amount of data that hurts you on your bottom line because you're storing it too many places. Yeah, so this you have to be careful, right? So I talked to a guy at Google, for example, a few years ago, we're saying for Google, they store everything they can about everything you type, everything you search. So it's a strategy for them to store everything. And I don't think like on the banking side, that's not what I'm saying and I'm correct with you, because if not, the cost will be way too high. Right. So I did a lot of work in the foreign exchange markets. in foreign exchange market, if you have like hundreds of dealers who make prices if you want to buy your exchange euro against dollar or British pound against euro and those price are updated every microsecond or millisecond If you try to ingest all those data point from all the producer of data you are going to fill a data room like very quickly and Alright, so you you have to be able to select what problem you want to solve Stephane (54:23.626) And for that, which amount of data you need to All right. And I think you are pretty correct there. Don't store everything because you, you it's like those old people who collect everything in their backyard and then pile up. You have to do what I think is relevant to you and what will solve a problem for you. But, and that's why it's good to step back, have that vision and that vision helps you to have a great data model of what's problem is going to solve for you and then brings that value back to you. I'd like to wrap up the discussion with Jani. There are no episodes in this podcast where we don't touch Jani. Probably that's happening across everybody in the media. So you've seen how industry is adapting and changing and the new fintechs are emerging. In your opinion, are there any critical blind spots that fintechs and banks are completely ignoring right now? Yeah, I think it's an opportunity and a challenge at the same time on GNI for banking. And I'm saying, my experience is more investment banking and not retail banking. So I will speak about that part of investment banking is it's very hard at this point to create a business case of how much revenue I will get from a GNI solution versus investments that I need to make. Stephane (56:08.014) And that's why a lot of investment bank struggle like crazy to launch GNI projects, because that's the first question from their management. will get one. You want some money or budget to do something. What is he going to do for me? Right. Like how much more revenue I'm going to create. And we don't have these answers today. And that's a challenge, but I think that's a mistake that we make. because if we don't embark to that journey, somebody else did, right? So there was a quote. I was in a kind of a GenEI breakfast meeting a few weeks ago, and one of the speakers, know, people are afraid because they don't know about GenEI, what it can do or cannot do. And they say, well, is he going to replace me or replace my job? And he said, well, it's not... AI was going to replace your job. It's somebody using AI who is going to replace your job. Right. And we start to decide in investment banking right now. If you look at some of the high-fricency trading firm and its public, so that's I can quote them. And if you look at the profits that Jane Street makes in 2024 last year, compared to the profits of a big bank like Goldman Sachs or P1. you will see how far those high-frequency trading firms who are using AI and Gen.AI a lot are making compared to a traditional investment banking position. So I think that's why for me it's critical, critical for investment banks to enter that space quite quickly, even if they don't know yet. And that's why my journey in innovation was that is sometimes you don't know where you're going. but you know you have to get there. You have to do something around that. And you can pilot some use case and I will give you one or two for me. And I see from asset management community, for example, every bank sent research paper to their clients every single day. it's not, know, one research document could be 60 pages, right? So what's the... Stephane (58:27.584) American administration done last week. And you can write about it every single day for 60 pages. But they receive that. a large pension fund, a large asset manager, really receives that from their top 20 banks every single day. They cannot in jest read 60 page times 20 every day. But they can use GNI to translate that information to them into a more concise format and to try to generate some insight from it to say out of those 20 reports 18 of them think that the economy is going to go to a recession but two of them think it's not going to go because such a chart region. So this is extremely powerful technology but you have to pilot different use cases and some of them will give you great benefits and some others probably will say well it's too expensive from the benefits that we can have. It's interesting, I will check it after the recording. I'm just curious. It's interesting if there are any cases left from the time when the industrial world and world in general was switching over to electricity and how it was sold to business owners because back then, electricity was something like Gen.I for businesses now. How can you predict that in 100 years that will be so powerful and will power the whole manufacturing world, right? So I think at least for me, the analogy is like that. It's really hard to quantify innovation in a sense. cannot create financial model because you just don't have the precedent in particular industry. In all these trees of challenge people face in something in investment banking is we already have a lot of structured data So people say well I don't need the generic because all my data that I use is already structured in database And I have too many of them, and I don't know what it is But but Jenny hi, it's both way right it's you can go from structured data to unstructured data Yeah, so same thing we can have something generate a morning call report for you to say listen. This is all the Stephane (01:00:40.302) price move during the night in Asia and boom, you arrive in the morning and you can read that paragraph generated automatically. going from structure to unstructured, all the other way around, like the example of research going from unstructured to structured data. So I think people don't realize yet how powerful it is. And that's why I'm passionate about it because I think the finance industry is lagging behind other industry sector. on the adoption of Gen.i. And it's strange because it's an industry with lots of capital, with lots of power to be able to invest in things, but currently it's running behind. Well, on that, we're going to finish the discussion. Thank you very much, Stephane, for being a guest. Thank you, Alex. It was great to be with you. And for the listeners, thanks a lot for staying with us. if you enjoyed the conversation, don't forget to the like button, subscribe, leave the review on the podcast platform that you're listening. And see you next time. Bye. Stephane (01:01:47.502) Bye.