Taktile’s CEO Raised $190M Betting Most Bank Jobs Won't Exist in 10 Years

0:00A few months ago, a 300-pound bronze lobster appeared in front of the Wall Street Bowl in New York with no explanation. For a full day, the internet went crazy, no one could stop talking about it, and tens of millions saw it online. The next day, it was finally revealed that a fintech company called Tactile was behind this stunt. We didn't say it was us. And then we had 20 million views on Twitter. Everyone talked about it. Was it OpenAI? Was it OpenClaw? Who did it? Media was there. Fox was there. It was a huge thing. And the next day we had to claim it that it was us. My guest today is Mike Taro Wechmeier, CEO and co-founder of Tactile, an agentic decision platform that helps companies across financial services in the entire world automate some of their most critical processes.

0:59Tactile itself is an agentic decision platform, the decision engine, the data marketplace, and the agent platform. And combining those three things actually allows you to make better decisions for financial institutions. You rip out the heart of a bank, you put it on Tactile, but once that's done, you're becoming a very, very close partner for many years. Mike and his co-founder built Tactile from an initial failed product and 99 bank rejections into what is now one of the fastest growing fintechs in New York and a company that has just raised over$100 million led by Goldman Sachs. In this episode, we get into their pivot, how AI has actually changed things for banks and financial services.

1:47And of course, we'll talk about the lobster. There's not that many things that we can identify anymore where an AI agent cannot beat a human, which is incredible. If you go to a big bank in New York and they have 10 ,000 people in compliance, I would be more nervous if I'm a compliance officer because I might not be needed anymore if the agents will do more and more of the work. I think the AI bank of the future will be very lean.

2:18All right, Mike, Miguel, Mike. Welcome to Fintech Leaders. Welcome to our studio in the middle of New York City. Well, thanks for having me, Miguel. Great to be here. Good to see you again. And before we get into talking about tactile, because that is super interesting, I think it's also interesting to learn how you first got involved in financial services, because I understand it was very early, as early as, you know, maybe first grade elementary school. Yeah, yeah. No, it's actually like my grandpa, he sometimes dealt with stock a bit. So he always wanted to raise us, thinking about different engine markets.

3:01And yeah, he bought a few stocks for us when we were kids. And we had this house. I'm from Dortmund in Germany. We have a soccer club called Borussia Dortmund. That's probably the most thing people know about it. And we were playing on the streets with the other kids from the neighborhood. And then every once in a while, he came to his terrace and was screaming, Microsoft is up 5%. And no one had any clue what that meant, but we were cheering. We were loving it. And I got excited about the rising stock price early on, but that was probably the first touchpoint before. Then I became an engineer, so that went down away from the financial markets.

3:39And then with Tactile at all, it all came back. Did that stay with you, like the stock trading? Or was it something that happened early on and then you kind of forgot about it until college? More forgot about it. Yeah, I wouldn't say I got deeply, deeply involved in trying to go into that field. I like math. And so I became an engineer. And through engineering was math and statistics. And then machine learning engineer. and then the whole thing spins back to what I do now. Yeah. So you grew up in Germany. How was it growing up in Germany? Like, what's the school system like? How did you figure out that you like math?

4:24Did you have the resources to really go deep? Yeah. It's like, you know, Dortmund is more of a coal mining town, so there's not, you know, that many exciting things happening. We got our soccer club, Rossi Dortmund, you know, which I'm burning for, I've always burned for. but parents were physical therapists. They always told me and my siblings, do whatever you like to be happy. So there was no big pressure from our parents' house. And it's a public school system, free. Everything is for free for everyone in university. So it wasn't a public school. And they just let you find your own way. So it was a very good childhood.

5:02Most of the time I spent running. Seal a runner? Well, sprinter, actually. I actually don't like running, running. I was a pro athlete. I ran for the German national team, track and field, 100-meter dash. So that was probably eight times a week during the time. I was 14, 15 until I graduated high school. So that's probably the thing I did the most when I was a kid. I didn't know you had made it to the German national team. So was the Olympics ever? No, it wasn't good enough. You need to be top three in the country. I was more like top 20. It's not enough. But the German government gave me a scholarship to go to the U.S.

5:44So I spent high school in Utah. I lived with a Mormon host family, which was fantastic. Mormons? Mormons, yeah. Really, really, really good to me. And why did I send you to Utah? There's no alcohol, no drinking. So it's the best place for a young sprinter to spend high school in. So I spent a bit of time there, but then got injured a couple of times, which is not good as a sprinter. So I decided to focus more on academia by the end of the day. Speaking of academia, so you focus on math. What was your undergrad in? I was engineering. Engineering. And then statistics, doing my master's. Did you consider doing a PhD?

6:21Because you did your master's at Harvard. Yeah, I was. I actually had an offer to do it. Back in the day, my mother passed away, so I wanted to be home for a while. And then my professor said, look, why don't you just write your master thesis with me? Then you have a first paper. And if you like the work, because you don't get a PhD to have a PhD. You get a PhD five years at Harvard. You get a PhD to be in academia. I wrote that master thesis. Oh my gosh, I was clear. I should not. I don't want to do that for the rest of my life. I think I'm just, I'm more like the business building side of things.

6:59Being out there, being entrepreneurial. but the academic part, I think I could do it, but it was not really the thing that excited me. And then you never get to your genius zone if you don't love it. When did you start figuring out that, like, I like to build things, when you start realizing that starting a company was even a possibility? Probably later than one people think. Like, you always see those stories of entrepreneurs, you know, they're like, grow up, they're 12 years old, and they always knew that they would want to do it. I mean in university I founded a lot of like clubs and like we did like project management for enterprises so I always like wanted to do something and it was more about probably exciting people for a vision or something that I wanted to do I think I was not you know the Elon Musk I want to like put rockets to Mars type vision so I actually worked for three four years post post-graduation as a machine learning engineer for kind of like a Palantir-like company.

7:57In Germany? Yeah, I started in the US, but then I was mainly spending time in Germany. But it's not that I graduated college and I was like, okay, no, I got to build this big company. But then over time realized, look, there are too many problems that can be solved with the product that we're building now. But then I decided to do it. And then there's that spark to start to launch Stacktop with a different product than what we have today. I think you've shared in the past that this was a pivot, right? At what point do you realize like, hey, what we're building is not our future, feel like we need to pivot?

8:37Yeah. I mean, at some point in time, you do feel it. I mean, like, I was always like when we were in Y Combinator, our group partner, Michael Seibel, like he always told us you have to find product market fit. I was like, sure, but how do I know? And then he said, if you ever push the boulder up the hill, you know you don't have product market fit. If you chase a boulder down the hill, you know you have it. And the first two, three years, so we're building an ML Ops platform as machine learning engineers, horizontal, PLG led mode, developer first. a very technical product. And in 2020, when we started Tactile, there was nothing like that in the market.

9:17So it was not a bad idea. But then in 2021, 22, all the hyperscalers, they built the same product. AWS SageMaker, GCP Vertex, Databricks has it. And like, you could feel, back to your question, you could feel, it was too hard to sell it because there was too much competition at some point in time. That was the one thing. And the second thing is, I mean, we pitched it to 100 banks and 99 said no. So at some point in time, you know, you should rather, you know, listen to your customers and see what else is out there. And did the insight for your current product come from listening to your customers?

9:56It did because we did have customers like machine learning engineers, data scientists at, you know, fintechs and banks. So we were already in the space we're in right now. We were only focusing on the machine learning part. And then we had the credit and fraud teams come to us and say, wait, wait, can you also like build a no-code, low-code environment around it? Because I cannot code as a credit analyst, but I would like still to deploy rules around the ML model. Not every decision is purely ML-driven. If it's a mix between rules and non-rules, we said, yeah, sure, we can build it. And then they asked us, can you build a data marketplace?

10:28We need, you know, like bureau data. We can think about that. And over time, they kind of led us into the direction where we are right now. Where is that right now? Paint a picture for us. Maybe share about your product suite, the scale, and then we'll talk a bit more details of how you've gotten there. Yeah, so TechTent itself is an agentic decision platform. So we only focus on financial services. That's it. You can do everything. We learned that in the hard way from the first product. We focus on financial services, which is fintechs, banks, also insurance companies. It's a big new vertical that we're attacking.

11:10It's really successful. We'll talk about that later. And yeah, we have the decision engine, the data marketplace, and the Asian platform. And combining those three things actually allows you to make better decisions for financial institutions. Meaning we focus on the most mission-critical use cases. So we don't do back office or we don't do call center optimization. We look at what's important for a bank. There's onboarding, KYC, KYB. There's credit underwriting. There's transaction monitoring. Like all of those things we want to be the solution to run with. So, yeah, it's a big change for an organization.

11:48You rip out the heart of a bank. You put it on tactile. But once, you know, that's done, you're becoming very, very close partners for many years. i've seen you in action you know doing sales if you want to call it that way or like you know building relationships around the industry and and obviously you know how to do it you're very very genuine but you mentioned that at the beginning with your original product it was just wasn't working right so I'm sure there was not just the product but also with the pivot you got a better product to sell but also you learned how to sell right like what have you learned about that also you know related to that initially I'm sure your customers were smaller firms and then you've graduated to increasingly larger and larger ones yeah the important to understand to answer that question is the journey uh the first journey was really plg that like you have to be in in in like on reddit and talk about your product and um now in contrast now we do a top-down sell like we go to the c-suite of a bank yeah we build up trust so it's very very different i think there's product market fit but there's also like the founder product fit and i think you know max i'm a co-founder i I think we are more situated for the top-down motion.

13:15You know, build up trust, very knowledgeable, and become the AI transformation partner for a large institution instead of building a huge developer community, which we would need to do with the first product. And I think it was just not what we loved. So I think first it changed a lot towards what we do now. But I think in terms of learning, I think probably as you always are as a young entrepreneur, you're incredibly eager and going back to ripping out the heart of an organization you can't be eager you know you need to you need to be need to build up trust over time it's a long-term game and you know you need to be very kind of like a thought leader in the space in the first place and then the sell comes afterwards so like we'd rather like want to be um the the institution that helps actually to run ai in production and once you've proven that and your customers talk about it, then the rest of the customers come to you.

14:10And in the beginning, I was just calling up people and trying to sell the thing. But that's not exactly how enterprise sales work. How much have you relied or leaned in on your cap table, your angel investors, your VCs? Is it zero? Is it a little bit or is it a lot? I think it was the first 150 customers were fintechs, or, you know, like Index and like Tiger, Y Combinator. They were incredible partners because they funded, they funded all of them. Many of our angel investors did, so it was like very easy to use them for introductions and come in with the trust of the investor being on the same cap table.

14:51I think now it's different. We are building like this huge advisor network of industry heavy hitters, former chief credit officer at American Express, for a claim officer at a big insurance company. And that is more the people that can help us now. Also, Index is not Tiger or whatever, which funds. It's not that they are connected into all the C-levels of all the big enterprises. That's our job to do it now. And this advisory board has probably been the most effective way of getting in and also building trust very early on because someone introduces you who they've been knowing for 20, 30 years.

15:31how do you incentivize that advisory board um we are i mean like those are all very busy people meaning um everyone has like an equity retainer um because they just have to spend time to meeting their friends for coffee and like chatting to us and then um yeah in many cases we uh if we grow as an organization we try to incentivize more yeah if they're being successful of course sometimes advice it doesn't work out which is totally totally fine and then you just say thank you and you're done after the year but like the longer you go the more we try to stack up the incentives for them to also you know get something out of the journey yeah um let's talk about building the company uh what's your approach to hiring i mean yeah hiring i think hiring for me is um less of a task and more of a more of a mindset meaning it's something that you do day in and day out you know no matter where you are no matter what you do um i probably hired probably a lot of people at weddings if i if i'm at a wedding you know i do enjoy the wedding but i talk to people you know who they are what they do and um we're in brooklyn here right now you know i was at a big halloween party in brooklyn two years ago okay and a friend of mine look at those two people over there we should talk to them I was like yeah let's talk to them and ended up one of them you know she joined us afterwards what was her costume it was a cool costume and I think that's it's a mindset thing and I keep telling that my um we have a value in the leadership team you know mind the diamonds like it's everyone's job day in and day out to look out for the best people and there's hardly any conversation that I have with anyone that I meet where I don't ask by the end of the day, hey, who is the best person that you've worked with?

17:19Such a quick question to ask, but normally one person comes to mind, and that is a very strong reference. So yeah, that's, by the end of the day, we are in a people business, high value going back to what we do. Like, we need the best people to succeed. So I consider that to be one of our main, main jobs. Yeah, I remember, I mean, many, many founders, but just as an example, Eric Glyman from Ramp. I think I hosted him three years ago or so, and at that point, he told me he was spending 40 % of his time just on recruiting, right? I suspect it's similar for you. I think it is because what we do at Tactile, I mean, I run pretty much every final interview.

18:03Now, not every, every anymore because we're too many people now. We have to hire too many people. But the last interview at Tactile is a values and motivation interview, which is a pure culture fit interview. and I do not look at all at all the former evaluations of the team. I hire an engineer and even if it's, you know, I don't want to know how good that engineer was because it kind of like disturbs my evaluation and it just values and I try to like figure out who is that person. Like does that person fit to us? I always ask the same eight questions and after, you know, I probably hired a couple hundred people and interviewed a couple thousand.

18:39I think by now I know what the right, Like, it's not the right or wrong answer in many cases. I, like, ask very wide open questions, like, what's your biggest achievement? There's no right answer to it. But, you know. You learn a lot. But you learn a lot what's important to those people. We hired one, you know, incredible account executive, and he said, well, I was best man at eight weddings. I was like, wow, that's pretty sick. You're, like, 28. I'm like, he must be a pretty popular guy. I think that's good for you to be in a sales role. And some people talk about private things. You learn a lot about asking open questions just to put them into their value framework.

19:14An interview is also a two-way sales process, right? What are you selling to them? What kind of culture can they expect? What kind of environment? Yeah, I think what's probably personally very important to me, that once we succeed, everyone gets a lot out of it. I think that I give everyone that promise. You know, you had the PayPal mafia in earlier days, and I think now we have incredible people that founded their own companies. I want to be proud of everyone saying that was the best place to work for me in my career in a couple of years joining Tactile. So I think that's what I promised them.

19:51And I think what motivates me personally a lot to build this company, not for my investors or only the customers or me, but if we all succeed, we succeed big, you will also get It's essentially something out of it for your career and, of course, also financially. And thinking of the organizational structure of the company, maybe share a bit of how have you structured a company? Is there something unorthodox or unusual that you've done to create a more productive machine, I guess? I think from the org chart, it's not highly unusual. I think one thing which is special about TACTA is our forward deployed motion, like the way that, you know, of course, Palantir introduced it.

20:38But we are incredibly white glove with our customers. We fly in on the ground, out of the helicopter and land with 10 people, and we're not leaving until the job is done. So like this heavy FTE motion, I think, is a big part of the success at the company. And you also learn a lot for like what product to build. As I said earlier, the pivot we knew because our customers told us what it is because I was like with the customers all day long and trying to like create that same culture within uh within the company so if you ask our investors there's sometimes why there's so many FDs but I think it's the right strategy uh and and and that's also why I think you were telling me last year you were in 100 and we were traveling 100 plus days of the year oh yeah 100 142 flights last year 142 flights Yeah, yeah, yeah.

21:27So, I mean, we have customers in 27 countries. About the org chart, so we have the engineering product team is in Europe. It's in Berlin, London, Romania. Now we're opening an office in Dublin. And then, you know, I'm in New York in the headquarters. We just opened an office in Sao Paulo. So if we have a lot of customers everywhere, you got to be everywhere. So that's time constraining, but it's, look, you got to do what you got to do sometimes. I remember meeting you a couple of years ago and clearly your mindset was international from the get-go, the structure of the company is international.

22:06But also I remember you telling me like, oh, I have customers in this region and this other continent. And it sounded as it came very natural. I mean, I know, for example, Latin America, it's been a good region. You just opened the Sao Paulo office. how do you sell into these companies from a new york network because a lot of companies a lot of fintechs would love to get the the same results i mean we were like growing up in the in the in in covid times which you know was helped you to get meetings booked online so like i think the world has changed so it became much easier i think from us um we localize a lot by the end of the day So we have, you know, all of our customers, like in Mexico, like it's Spanish speaking.

22:52I want people to be there on the ground. The same for Brazil. Like you need to tropicalize. So I think we're just trying to understand the market really, really deeply. Of course, like that's 27 countries, but now we're into scaling mode. I think what sometimes the, it's not purely positive to be in so many countries. Right. In the beginning, there's always some low hanging fruits in each market. So that's where there's so many, but you can see now from the focus of the company, we do focus on a couple of core markets because you build up a reputation, you build up an office, you have like the key logos, lighthouse logos in the country.

23:25So I'm actually like, we're actually getting a bit closer now again to like really scale in some core markets. And so it was not, it's not only glory if you're everywhere. Do you think about it by cities, not even, not even countries like these are the hubs? More like similar cultural regions. So like, Brazil is Brazil. You got to be there. It's different. Most big banks are all in Sao Paulo. The next office will be Mexico City. And then you can serve Peru and you can serve Chile. And I think I'd rather think about cultural alignments. And then, of course, the market needs to be big enough. But right now, US, UK, Germany, France, they're all big enough markets for where we are.

24:13What was the sequence of launching products? Because now you have a full suite, but you started with one specific use case. Maybe share about that sequence. And how was your launching process? So we always wanted to be a platform which is hyperflexible and can serve many use cases. The cadence was, let's start with the most complex use case from a decisioning perspective, which is credit. Credit is highly complex. You have statistical models. You have decision tables. So we knew that if we would nail credit, it would be much easier to go to compliance KYC. KYC is not that complicated. And I think thinking about it from a technical perspective in the first place, that was the one thing.

24:57But then you can't always predict what's going to happen. I think even three, four years ago, you were building a product by talking to customers. They were telling you what to do. and then you went back to your product team that were building it. I think now with AI, it moves so incredibly fast that you do not know exactly what AI will be able to do and what use case they will unlock. So now you need to build a product under the assumption that everything will be possible. And that's different than four years ago. I could say the most complex use case, the other use case, I could kind of like make a plan.

25:35And now you need to think about from a product perspective very differently, which makes it a lot harder. So now it's different. We're just trying to operate under the assumption everything will be possible with AI. And then we do think about rather the biggest use cases, not what customers tell us what they want to do or what's possible now. So that's changed in this world a little bit. You're saying will be possible. So today we're still not there yet, but a lot is possible, right? When it comes to what you've learned, not just your product, but products you're seeing in the market, maybe even outside of financial services, like what has blown your mind in terms of what is possible today?

26:16Yeah, I think probably the big one was over Christmas last year, you know, like Opus 4.5. And it's like, I mean, like we've been trying to do, using agents to do financial spreading for SMB underwriting for like three years since the LLMs came out. I could see like other verticals to be disrupted. I could see like legal being disrupted with AI with like a Harvey. It was like, LNM was really good at reading text and helping lawyers. And then I could see it with, um, cause and optimization, like a Sierra and Decagon were like, okay. And I was like, okay, that's possible. I want to rebook my flight at United Airlines.

26:53Okay. It's not, it's possible, but it didn't work for us because we are not only living in a world of text and speech. We're living in the world of numbers and structured data. and you were giving like a financial statement and a balance sheet and a cash flow statement to the agent to do financial spreading, it's never, ever worked. Nothing came out of it. And then, you know, we keep doing it, you know, once a month. And then over Christmas, we did it again. Unbelievable. Like this stuff just all of a sudden worked. And I think that's probably been the biggest acceleration for us over the last, you know, six, nine months of now building agents that actually operate better at complex tasks than humans for the use case that we care about.

27:40And it's very exciting to see. I think there was, I'm trying to pull out the numbers, but you had a test where the performance, humans versus machine, machine was like 90-something percent and humans was low 80s. Yeah, there was also financial spreading. Financial spreading. Yeah, exactly. And so if that's the case, right, something we've heard for a long time in the industry is this process is with a human in the loop, which in my mind has always been twofold. Number one, you needed the human, so the result will be actually what the client wants. but also it gives comfort to highly regulated institutions.

28:32I'm sure there's other reasons.

28:36What remains human in the loop and where are humans completely irrelevant? I think you should always have the possibility to have the human in the loop. And out of different, like every use case has a different reason why. Sometimes it's like, because you don't want to make a big mistake because it costs you a lot of money as a bank. I think if I give out a loan to an SMB and it defaults, it might cost you$2 million. For AML, the regulator really wants you to have a human to really look at something. So it differs a bit. But I think the way that we think about it and the way that we build the product is just combining the two worlds of decision engine and agentic decision making together.

29:21And if you look at, I go back to that example, because we talked about SMB underwriting. even though like financial spreading is really good they're good at the financial spreading piece which is not the underwriting decision it's just getting all the documents in the same form the regulator doesn't care about the pre-analysis about reading out documents perfect for agents but then the final credit decision you still want to have your credit policy and your decision engine to do it so despite like having these interesting evas and benchmarks where we see that agents perform better than humans at certain tasks and i would always think about tasks rather than end-to-end decisions.

29:56Because in the end-to-end workflow, you want to have both. Sometimes you want to automate a decision engine. Sometimes you want to have the human in the loop and the case manager to take a look at. And being able to mix that, I think that's a key to success. And then, back to what I said earlier, the models will even get better and better, but having the possibility then to shift more and more of the work to the AI agents is something that you want to be able to do, even though they might not be able to do it today. So it's a dynamic optimization problem. And you definitely still need the human.

30:29But look, in three, four years, potentially not anymore. Yeah, if you think everything will be possible, what's that future that you're imagining in three, four years? Yeah, I think there's, I mean, there's one question is what will, you know, agents be able to do. But there's more to it. there's also a cost question. You know, like people are being very proud sometimes of using a lot of tokens. Not anymore. I mean, look, if someone shows me, if another founder shows me again, they're bored of like, who uses the most tokens in the organization? Like, I don't think it's a great metric. I mean, because not all decisions in financial services are made for LLMs.

31:15Sometimes you just want to take rules. You want to put them into code. and that is a marginal cost of almost zero to run it. Some of our customers, they were like, I want to identify everything. I was like, even if it would be possible, what you asked, what would be possible? It might be able that everything is possible with agents, but sometimes you don't need it. And I think knowing where the technology, like large language models can do things that humans could do before, that's a good proxy. But if you want to use an LLM for a decision, which just, it's perfect and good enough because it's a rule-based system and it just would be so much more expensive to let an agent do it even if the agent could do it.

31:55I think like that, that is, I think one needs to like have in mind. Something that I've really liked, I was having a conversation with Luis, who's the founder of Cloudwalk, Luis Silva. Amazing. Oh yeah, amazing. Brazil. Yeah. I was in the office. Oh, crazy office. They have a basketball court in the office. They have a huge flag of a capital, I saw that, yeah. And he was saying, he was one of the first ones years ago to say to me, like, AI is going to make us more human, you know, because it's going to empower, actually, the more human nature, right? The creativity, the face-to-face interactions.

32:39But there's companies, and I don't work at Facebook. I don't have too much insider knowledge. But reports are that morale in many teams on Facebook is pretty low because of their approach to try to agentify everything and even track the keystrokes of everyone. And so it sounds like it's becoming clear that there's a right approach, there's a wrong approach. Have you felt some of this in your organization? Is there more excitement? Is there also some people getting nervous? How are you dealing with this? I think for us, right now, there's still more excitement because AI makes all of our people more productive, but it will not replace them.

33:28I think if you go to a big bank in New York and they have 10 ,000 people in compliance, I would be more nervous if I'm a compliance officer because I might not be needed anymore if the agents will do more and more of the work. I think for us as a tech company, of course, engineers love it. They code a lot faster. HR people is loving it because they don't have to, you know, send out individual emails on LinkedIn anymore to hire someone. It's all automated. So for us, I think it takes the tedious tasks away. It makes people more productive for the moment. So I think it's rather a positive sentiment for now.

34:03But, I mean, we try, of course, always to put pressure on everyone in the organization to use a lot more AI. It is part of the performance feedback attack time by now. We don't have a scoreboard for token use. I think that's dumb. But we have every manager measuring on how much AI has been used by whom. And if not, it has a negative influence on your performance feedback. So some nervousness maybe if you're not ready for it. But I mean, everyone can do it by now. Right. You're either being, you know, you don't want to do it, but you can do it. Yeah, yeah, yeah, yeah. there's a lot of bears also out there for AI.

34:44Let's give them a little bit of a can of fodder, a little bit of food. What are some of the aspects of AI that you think are still overhyped? I think it actually is more than the point I made earlier that you just don't want to identify everything in financial services. I think people are, many of our customers to say, like, forget the old world. Everything needs to be Asian. It's just not true. Because, like, many people just don't understand what... Asians is not always a net positive thing. You know, and I think thinking just always have a good rule, like, where is it better? But also, what's the cost that analyzes it?

35:23And what's your dependency as an organization? Especially now in Europe or European customers. I mean, export controls for LLMs. Like, we're not in a good position in Europe to play the big game. with AI. It's all here in the US. And I think that's, and then if a manager of a big bank in Europe tells me everything he's agendapied, I'm like, okay. But like, you should also think about like, you know, how do you localize? What do you still own? How can you use open source? How can you host your own model, which is cheaper and you're not so like dependent on the big lab? So yeah, only more, more, more agentic.

36:00It's not nuanced enough in the world that we live in right How does your personal AI stack look like? I mean, for me, everything is like code code in the meantime. And that's pretty much my productivity suite. I mean, we use it. I touch it in all of the HR tools and the performance tools. It's in everywhere. But for me personally, I don't code that much anymore. So it makes things a bit easier from a tech perspective for me. That's where a lot of the cool tools are. What I keep thinking is that as a non-technical professional, if you just master Claude and spend a lot of time with Claude code and co-work and also everything else, you're already 80, 90 % of the way there.

36:47Yeah. I mean, you just need to know how to do it and what you use it for. But even for my, my role is, has a lot of heterogeneity in there, meaning mainly I need to get a lot of information on a lot of things. That's what LNs are really good at. You know, you connect it to your G drive, you connect it to your Slack, you connect it to Gong, you connect it to HubSpot. By now, my chief of staff and I, we've built a system and a tool that gives me full overview over every customer interaction, all the top deals, when we talk to them, how we talk to them, what was the sentiment, where I'm needed. It's unbelievable.

37:25That had cost him and me normally hours and hours beginning of the week to prepare for the week. Even having an overview and oversight of what's going on in an organization, it's unbelievable as a CEO. You just have full transparency. And then you can, my main job is to make decisions. And I can make a lot more decisions a lot faster because I have so many more information at the place that I needed in. And importantly, you're not working less. You're working more. Definitely more effective.

37:56So let's talk about your branding. You pulled a really good marketing stunt. Tell us about it. Yeah, I mean, we were launching Tectile Labs, which is our research institute that has the task of building evils and benchmarks for the use case we care about. OpenAI, they do evils and benchmarks for how good is AI to pass the bar exam or the medical test, which are very interesting evils and benchmarks. But our customers and we care about how good is AI at SMB underwriting or at KFC. And there was nothing like that in the market. So we said, okay, let's build tactile labs, have a research council with industry leaders and experts on it and give the industry what they need.

38:46That was the first idea. My brain doesn't work in the way that a great marketing brain should work like. So I thought, let's just put tactile labs launches at a brickstone wall in Soho or something. Then I talked to a friend, Nico, who runs a creative agency. And he's like, no, no, no, wait. We should think about it in a different way. and at that point in time we just had Open Claw which was quite open AI and Open Claw was like being criticized for the security breaches potential security breaches is AI safe or not what we do is bring AI to banks in a safe manner okay so you had that and what stood for like the lobster was actually the sign for Open Claw so we thought we just let an artist produce a 300 pounds lobster.

39:34And we put that lobster in front of Wall Street, in front of the bull. And then you can see AI incorporated in the lobster, being persecuted for no security, fighting with Wall Street and the bull against each other. And we just put it there in the morning, no permit. We just thought what's going to happen. And then the whole internet, we didn't say it was us. And then we had 20 million views on Twitter. Everyone talked about it. Was it OpenAI? Was it OpenClaw? Who did Media was there. Fox was there. And it was a huge thing. And the next day we had to claim it, that it was us. But I think it really sparked the discussion in the industry, how ready is financial services for AI?

40:17And the answer you find at Tactile Apps. Because we do the EVA's benchmarks to give an answer to managers and bank executives to see what works and what doesn't work. Genius move. I mean, sounds like the impact views-wise and sounds like I'm sure engagement, I'm sure you got new inbound. Sounds like it was very, very positive. I saw it and then I was very glad to see it was you guys. But because I was lazy, I didn't walk down to see it. I should have. Where is the lobster now? It's in the office. Oh, nice. Yeah, no one took it. It was pretty heavy. No one took it. And by night, we took it back and it's in the office now and people like it.

41:00It was a cool moment for the company. That's awesome. I need to stop by to see it. Yeah, always invited. And so tell us about some of these benchmarks. Are banks ready? I think more and more, I think for most things that we are testing for right now, we just launched another agent on Adverse Media Search, which is a part of the KYB process. And there's not that many things that we can identify anymore. where an AI agent cannot beat a human, which is incredible. So the message is the technology is there. But I think what we are, why is not every bank just coming to Tactile and buying all the agents?

41:39Because there's still a big trust gap in the industry. Banks are regulated. You don't want to make mistakes. So I think a lot of the education that we are doing right now, a lot of the conversations, back to your question earlier about the sales process. It's not so much about, oh, look how cool Tactile is. It's more about, let's talk about what's possible and let's show you with data on how good the agents are. And then we can think about what's the right point in time for you to roll them out eventually. It's not all smooth sailing, building a company. I know it's very tough. Can you recall what was maybe the toughest moment building Tactile?

42:23um i think definitely the point in time where when we realized that the first product the mlb product is not really working but you have once you have a team and they build it and that was invested into it then kind of like making the decision to throw 200 lines of code into the trash and rebuild because it's no it's not always so crystal clear like you have some customers they're paying a bit but is it good enough is it not good enough so i think like making this making that decision to really cut what we had sit down and rebuild the full thing i think it's you know no ex post that's like it was the best thing but that point in time was like pretty scary but fortunately we never had the moment where we had to let people go or like some big customer I think we had the smooth ride so far.

43:17Yeah. But that was probably difficult. I was going to ask, did some engineers or team members, did they rebel? No, I think by the end of the day, people know if something doesn't perfectly work, and they're all there because they're so proud of what they've built, because they want to, back to my point about it earlier, hiring based on culture. I try to hire people that want to achieve something great, and no one can be happy by building a product that no one wants to buy. Yeah. How about your investors? I mean, how did your investors react? I think they were also bought in. I think, of course, they don't have all the data points.

43:58Normally, you know, you talk to your investors, you're trying to paint a somehow positive picture for like the day-to-day toughness of selling it. I think they went on there, so it took a while to show them data, show them conversion rates and say like, The market has just turned from 2020 to 2023 when we made the pivot. The market is different. What you invested into was a great market opportunity two years ago, but it's just not anymore. I can tell you five hyperscaler products which are similar. Do you want to build a big business or you don't want to build a big business? Nothing. The investors we have, they're very, very ambitious.

44:33So they weren't born fast. Took a bit of education. But at the end of the day, a good investor should trust you because no one's closer. to the action than you are. Exactly. Like, I'm doing this, you know, because I think we can build a great company. And if I'm convinced that the best way would be, I'd rather miss a lot of things or get the wrong judgment. So when it comes to fundraising, your latest run was quite big, 120 million. Was it Goldman led? Goldman led, that's right. Yeah. But you've had, obviously, previous rounds. What have you learned in fundraising and maybe what are some of the mistakes that you made maybe some early stage founders can avoid?

45:21I think my wife used to be a venture capitalist. So I had a lot of conversations I listened to over the years of understanding the psychology of investors. And sometimes they're like little kids. It's a formal business. like there's one kid in the sandbox and then once there's one toy and then everyone wants to have the toy. So I think like understanding that it's not only about the product and the revenue, the dynamic of a fundraise happens just by the FOMO and demand within the VC community. And it's pretty small community. There are not that many relevant people that can invest into. There'd be 50 in the world that are relevant.

46:01You know, it's like pretty small. They all talk. And they all know each other. The main currency of, you know, of like junior VCs is information. So once someone knows something, everyone knows something. So understanding the dynamic of that ecosystem, I think is really important for a good fundraise. Always product first, customers first, revenue first, but only because you have that, you can still have the same numbers, a really bad fundraise or a really good fundraise. I think I understood that early and I always try to be really mindful on how to structure such a process. Like when do you give whom information?

46:35How much time do you give people, you know, to look at your data and make an offer? Do you have like some VCs that might like you so much that they give you an early term sheet that accelerates just around dynamics, but don't give anyone too much time? Don't try to spend time with all investors the whole time. In the beginning, I thought, oh, if I'm best friends with all the investors, they're all going to give me money. Well, look, they're all pros. They want to make a good investment. It doesn't matter how much they like you. So I think by the end of the day, just understanding the dynamics of the ecosystem and then running a structured process with a couple of stalking horses in the game.

47:10Let someone run fast and then squeeze it. And then I think you can influence the outcome of the fundraise a lot. Does that remain true in your experience in later rounds? Not so much. I think now VCs become more bankers than believers. you gotta have the numbers yeah now you now you gotta have the numbers and of course there's always a dynamic in every fundraise um but i think now series d and then you know series d next year i mean like it's edited here it's it's a numbers game by now as well yeah yeah if you think about kind of the the next couple of years for for tactile you know what was what's like one thing that has you excited?

47:58For me, actually, breaking into insurance is huge. If you think about like first principles where AI is good at, AI is really good at automating things, you know, reading our documents, like those type of tasks, you have a lot of that in insurance. You look at claims in insurance, it's, you know, it's such a tedious process. and we've just seen tremendous success with one of the biggest insurers in the world runs a substantial part of their claims on tactile. And the business case behind it is so huge. So I think I'm very, very excited about insurance and we're investing a lot into that at the moment.

48:40And I think a big part of the business will also come from that, a big growth part of the business. All right, so we're going to do some rapid fire questions. All right, let's do it. who's a leader in business or outside of business who's a leader you admire and why yeah i mean probably sounds you know very uh cookie cutter but i mean like jamie diamond but not because i don't know how good of a leader he actually is in day-to-day culture and everything he's building i think what he's just a big part of my job is you know representing the company You're almost like a running billboard in the world.

49:17And I think he's just done such a tremendous job of getting respect in the ecosystem. Every time you see him anywhere, the way he talks, the way his thoughts are like structured, the point of view that he has on topics, I think it's just extraordinary. Almost like, you know, a white shirt blank vest, I don't know how you say it, but like, I think that's what I really respect about him. But yeah, I don't know how he runs this company from the inside. You get 10 minutes with him tomorrow. What do you ask him? I'd probably ask him I'd probably ask him like honestly how much does he prepare for all those things?

49:51Because like I do sometimes I prepare a lot for like an interview or something. Sometimes I don't do it at all and I think like for him everything is always like on point 10 out of 10. I would actually like to understand how much time he spends on doing that or it comes very natural because it's two different strategies and I actually run better not preparing than preparing but I also don't get asked that hard questions yet. Well, I need to step it up. Well, actually, you and I saw him together at the Acquired recording. Yes, that's right. That's so true. That was a lot of fun. All right, so all these books, maybe they've been written by a guest or recommended by a guest.

50:37Do you have one or two books that you often either reread or recommend? Generally, I love biographies. Number one books that I read. There's Titan, Rockefeller. It's like 700, 800 pages. I haven't reread it. I just read it once. But it's like, what a journey. What an entrepreneur. Shaping the industry in the early days in the US. I think it's a really fun book to read. But as an entrepreneur, you learn hardship. That's a good one. That's a good one. I'm visiting Germany this weekend what do you recommend? you want to have fun or you want to chill? I got both versions in Germany let's hear the fun one you have to be in Berlin and you have to stay up a bit later to enjoy the recommendation but they are very fun techno clubs in Berlin and there's still a thriving techno scene in Berlin No, no, no.

51:40The world comes to see that. I give the record, the details of the record. I've heard some crazy stories. I'm sure they're all true.

51:52What's a project you're working on right now that has nothing to do with tactile? Fan in the world. I have a sauna in my apartment in Berlin. Trying to get one in New York. Very hard with the permissions here. But I actually want to build a B2B sauna business in the US. I don't have time for that right now, but, you know, everyone's like, saunas are popping up left and right, other ship and like bathhouse. And that's the consumer versions of it, but someone needs to equip them with the saunas. Like every big hotel wants to have a sauna. I've, you know, I'm pretty far down the line. If there's like any entrepreneur who wants to build a sauna business, I need someone who can do it for me because I don't have time.

52:27But I'm actually thinking a lot about that right now. I'm on board with this. I mean, I've been going to the same chain of gyms for a long time and 10 years ago the saunas were empty now they're always full right that's a huge shift um but i was just in amati kazakhstan 10 days ago and i went to like the the typical russian kazakh sauna that i'm sure you've been to yeah i was there yeah i was there for like two hours full experience oh man it's unbelievable now that's where i get my sanity back yeah yeah exactly Exactly. And you can't do anything but that. Right? I hate people that go into the sauna with a phone or they're talking loudly.

53:12No, no. It needs to be quiet. It needs to be quiet. Okay. Good. Good, man. And then how do you think banks and tactile will look in 10 years? I think, yeah, I get that question I get asked by, you know, bank executives. I think the AI bank of the future will be very lean. A lot of the things that the bank does doesn't require humans. Just because of regulations and the financial crisis, like there was only one way to deal with the mistakes that were happening in the past is by, you know, putting bodies on the problem and there was no technology to do it. Now, as we talked about today, technology is there now.

53:57Visit the AI agents to do a lot of the work. So it will be very, very lean. I'm not talking about wealth management, high net worth individuals where you need the relationship. I'm talking about the day-to-day operations that a bank runs and it will be a very lean organization. It's also needed. Banks are bloated. If you look at the margins at banks, it's not what it should be. But they have the opportunity now to do it and to change it. Rob Haver from Motiv or Motiv I think that's how they say it. He was saying banks should have 5 % of the people. I do agree. Mike, thank you for doing this. Thank you.

54:33I really admire you and appreciate you stopping by. Thanks so much for having me. Thank you. Thanks for tuning in and I hope you enjoyed this great episode with Mike from Taktow. If you want more episodes, make sure to subscribe, follow, and leave a review on Apple, Spotify, or YouTube. It helps and means a lot. and if you have any suggestions or thoughts about the show just drop me a line on LinkedIn. See you next time.