00:41:12 · Guillaume Moubeche
[00:41:12] S1: And why did you like chose DeepL? Because you had like a lot of choices, I assume, you know, like you... You've worked at amazing companies, did IPOs, etcetera. So why did you decide to to move to to DeepL?
00:41:26 · Martino Cadoni
[00:41:26] S0: Yeah. I think several reasons. Right? First of all, is the sector. Right? So when I was I was definitely too young when Internet came up, right, in the early two thousands, then I missed the mobile revolution because I was in in banking and AG back then. Now it's the really the era of AI. So I say, I don't wanna miss this train. Right? I'm young enough, and, really, I wanna be an active part of this new revolution. I wanna be in the driving seat. So the fact that DeepL is a global AI company built in Europe was definitely a big incentive for me to join. The second one is Jarek. Right? I... I'm very privileged because I... Alongside... Yeah. In my career, I worked alongside amazing leaders, both CEOs, CFOs, founders. And and Jarek is is an amazing leader. Right? He's obviously is a scientist, so deep thinker, technical founder, but also most importantly, is is a great listener. Right? He's very humble, down to earth, and he really is a is a great leader that can really trust others. And he has built a fantastic team around him. So being part of this team, I I work from him was definitely another big plus. And then another reason was the cap table. Right? Having a... Being a CFO in a company where you have in the cap table investor like Benchmark, Index Ventures, IVP, Iconic, B2Venture, GIC, Atomico, Ontario Teachers is like... Is a cap table second to none. So the opportunity to work alongside these great investors, learning from the best, from legends in the industry like Matt Cohler and Danny Rimer and and others, Again, fantastic opportunity. And then last but not least, I think... I truly think, you know, there's a huge opportunity in AI. There's a huge opportunity in Europe. So being part of of a company which is still in the middle of this journey, it was like, you know, an opportunity I couldn't say no to.
00:44:09 · Guillaume Moubeche
[00:44:09] S1: Super interesting. And I'm quite curious because I think I heard you like discuss also about the the costs related to to AI. And we're seeing this, you know, with... I mean, there's a lot of controversy around, like, the models, like, requiring more and more compute. Like, some people say that they will never be able to become profitable, etcetera, etcetera. Like, you've been in companies that were like banking and based on software, and now you're like in the AI company. Can you maybe like give the difference in term of cost and how exactly do you like balance the the cost of AI? And what's the long term vision? And how do you make it align with the the financial prediction basically?
00:44:53 · Martino Cadoni
[00:44:53] S0: Yeah. Only from a financial perspective, then the business model is quite different. Right? And and I will say credit to fintech. I will say as a CFO, as financial operations, I think in term of complexity, sophistication, nothing beats financial service and payments. The complexity there is huge. And it also comes from the regulations and all the constraints you have. Right? Imagine if you are a bank or a fintech, the volume of transactions is in the order of millions every day. And if something goes wrong, you are in big trouble with, you know, regulators and stuff. Right? So I think from a financial operations, I I would say a fintech and financial services is more complex. But what is more complex in AI is really the uncertainty about the business model. Right? Like like, for example, think about billing. Right? Now there is a big debate in the industry, and nobody has an answer yet. What's the best billing revenue model in AI? Is it the traditional subscription per seat model. People is now try... Yeah. Starting to think maybe it's not the best for the high era. Then you have usage based billing, which a lot of companies, including us, are adopting. And then now it's like people shifting towards outcome based billing. Right? Can you base... Can you bill based out of an outcome? Right? So this uncertainty in the business model makes very difficult also to forecast because how can I forecast my revenue in the next five years if we don't even know in the industry how everyone is gonna bill the customers? If it's per seat, per outcome, per u... Yeah. User based. So there is a lot of variables coming into play. Right? The second one is, like, the CapEx. Right? As you said, the big LLMs are spending billions and billions in in development and training these models. Here, there is a very fascinating debate going on in the industry. Some people say, you know, those companies... Those three companies are playing a different game. And, you know, they have so much funding, and they're gonna lead, like, that technological advancement. And then all the others will have proprietary models and and specialize on something rather than having horizontal models. Others, very nonconsensus investors, they think, hey.
00:47:55 · Martino Cadoni, Guillaume Moubeche
Maybe we are too early in the industry so that something very severe is gonna happen at some point, and then there's gonna be a reset like we had in the dot com bubble. Right? I hope it's not the case, but I've told... Yeah. I've spoken to some investors who think that could be the case. Right? [00:48:20] S1: What do you think it's like... What do you think it's the biggest difference between like the dot com bubble and what we're leaving now? [00:48:30] S0: I think, I mean, personally, I see now the impacts on the ways of working and on the business models seems to be much more clearly defined. They also... The ability of AI to be applicable to substantially everything in much faster way. I mean, I think it's clear. AI is here. It's here to stay. It's clear. It can make a huge impact on productivity. While in the .com, obviously, was... The fragmentation of things that were popping up was, I think, a less of a clear use case. But, again, I was maybe too young to compare because, yeah, back then to have a proper comparison of the two. But when it comes to AI, I'm very convinced. I mean, this is is here to stay as there is not a bubble from a technological point of view. If there is a bubble from a valuation point of view, that's a different story. But I think that technology is real, and the technology is you have to stay. Right? And and I think those companies, you know, and DeepL as well, like, the charge and disrupting those industry from within, it is something which will have a lasting impact. And I don't see what what whatever can stop this. Right?
00:48:30 · Martino Cadoni, Guillaume Moubeche
[00:48:30] S0: I think, I mean, personally, I see now the impacts on the ways of working and on the business models seems to be much more clearly defined. They also... The ability of AI to be applicable to substantially everything in much faster way. I mean, I think it's clear. AI is here. It's here to stay. It's clear. It can make a huge impact on productivity. While in the .com, obviously, was... The fragmentation of things that were popping up was, I think, a less of a clear use case. But, again, I was maybe too young to compare because, yeah, back then to have a proper comparison of the two. But when it comes to AI, I'm very convinced. I mean, this is is here to stay as there is not a bubble from a technological point of view. If there is a bubble from a valuation point of view, that's a different story. But I think that technology is real, and the technology is you have to stay. Right? And and I think those companies, you know, and DeepL as well, like, the charge and disrupting those industry from within, it is something which will have a lasting impact. And I don't see what what whatever can stop this. Right? [00:50:06] S1: And what do you think are, like, the the biggest threats to to depot?