The Exchanges

Q Did they have the idea for Lagor or Lea at the time then?

A The idea for the business was not a product. It was a problem space. It was the fact that AI and legal is going to be a thing. And they had been playing around with the early BERT models that came up from Google, and also a, a version of them called SWE BERT, They were so incredibly bad, frankly. I mean, you couldn't really solve anything back then, and so they had built their own document management system, they had built their own editor, they had built, like, a legal research platform, they had all this code, and when I came in, I sort of said, let's just delete all of it, and let's build on GPT, and let's, like, build a version of it that, at the end of the day, lawyers are benefiting from.

AI assessment note: “The idea for the business was not a product. It was a problem space.”

Q Before I grill the shit out of you, uh, 60 seconds, what does Legora do just to set the scene for people that don't know?

A Yeah. Legora is the platform where legal work happens. I think that's a better pitch than the one that I had last time. And what started to happen more and more is AI is doing more and more parts of legal work, and this has to happen on a centralized platform. And what we started out with was I mean, simple assistant-based use cases, but this has grown tremendously, and it's solving different types of tasks for different types of lawyers. So if you are a transactional lawyer, and as part of a due diligence process, you need to review the data room, and then you need to find the red flags in that data, Lagora can do it. If you are a litigator, and you are preparing a brief, and you are drafting that in Word, Lagora can help you do it, right? So more and more of these tasks are being Uh, bundled into the platform. And what we're seeing is that, um, a bigger and bigger part of a lawyer's day is being spent on Legora, which is amazing. That's my favorite data point.

AI assessment note: “Legora is the platform where legal work happens.”

Q When people think AI law, there's a ton of fucking players around the space and around the verticals, but There's you and there's Harvey. And if we're blunt, Harvey is the first name that comes up. When you think about that, why is that?

A I don't necessarily think that's the case anymore. And the reason I say that is, um, I just saw this report. I was on Bloomberg a couple of weeks back and they had a big infographic that said that the most deployed generative AI tool in the top 200 law firms in the UK Outside of Microsoft Copilot is Legora. Number two was Harvey. We are moving, and the category is moving at such a rapid pace that it doesn't really matter who was first. It matters who's best, and it matters who the clients actually are coming back to and want to do more work with. So what often happens, right, is The firms will throw many vendors into a bake-off because they're in this kind of luxury position. I mean, basically they're playing VC. They get to bring in all these different vendors and they say, we're going to do a bake-off. And in the bake-off, it's up to the vendor to display why you are their partner of choice. And I say partner of choice because I don't think that these law firms or big in-house legal teams are Buying just a solution. They are buying an outcome today, but they're also buying, um, like an outcome tomorrow and they're buying into a vision of what an AI enabled legal team can look like. And so when they look across the board and they see all these different companies, they make qualified bets and more and more and more, um, seeing those firms make that bet on Legora. I mean, this …

AI assessment note: “I don't necessarily think that's the case anymore. And the reason I say that is”

Q When people think AI law, there's a ton of fucking players around the space and around the verticals, but There's you and there's Harvey. And if we're blunt, Harvey is the first name that comes up. When you think about that, why is that?

A I don't necessarily think that's the case anymore. And the reason I say that is, um, I just saw this report. I was on Bloomberg a couple of weeks back and they had a big infographic that said that the most deployed generative AI tool in the top 200 law firms in the UK Outside of Microsoft Copilot is Legora. Number two was Harvey. We are moving, and the category is moving at such a rapid pace that it doesn't really matter who was first. It matters who's best, and it matters who the clients actually are coming back to and want to do more work with. So what often happens, right, is The firms will throw many vendors into a bake-off because they're in this kind of luxury position. I mean, basically they're playing VC. They get to bring in all these different vendors and they say, we're going to do a bake-off. And in the bake-off, it's up to the vendor to display why you are their partner of choice. And I say partner of choice because I don't think that these law firms or big in-house legal teams are Buying just a solution. They are buying an outcome today, but they're also buying, um, like an outcome tomorrow and they're buying into a vision of what an AI enabled legal team can look like. And so when they look across the board and they see all these different companies, they make qualified bets and more and more and more, um, seeing those firms make that bet on Legora. I mean, this …

AI assessment note: “I don't necessarily think that's the case anymore. And the reason I say that”

Q You said about like partnership and that being very central to how they think and how they choose. You know, I had, um, Matt Fitzpatrick, the founder of, or the CEO of invisible, which is kind of like a McCore or a Turing competitor. And he said that essentially it is impossible to sell into enterprise without an FDE model. Do you agree with that? And are you seeing that?

A We have a very big A team of legal engineers who are ex-practicing lawyers from the top tier best firms, but they're not fully seconded, right? They're forward deployed in the sense that their main job is to make you successful. An example would be, um, a big firm that just went with Legora Widencase. Wide and Case now has the challenge of adopting AI across their entire firm. It's a big firm, such an enormous change management undertaking to equip all the lawyers across all the different practice areas, across all the offices, and across all the skill levels from associate, senior, associate to partner with AI proficiency. And we need to make them successful because if they are not successful on Nagora, um, you know, A year from now, two years from now, it's going to be a really sad conversation. So we invest a ton of upfront manual labor and time and effort in doing the implementation and activation right. So I do think that's necessary for enterprises where you're changing the way they work. If you just think about a process, right? Like let's say you're working in Um, for if we're just staying in legal, if you're working with AI contracting, you basically have a contract lifecycle management system. You just send a document somewhere, generate some red lines, and then you put it back. That's pretty easy. You don't need a forward deployed engineering or legal engineering mod…

AI assessment note: “So I do think that's necessary for enterprises where you're changing the way they work.”

Q What do you know now that you wish you'd known at the start of Leia?

A What the intensity of doing and thinking about nothing else would, um, sort of impact your own psyche and personality. For the last two and a half years, I've basically done nothing else than to think about Leia or Legora or the business, and You know, when you're in, when you're in college, or prior to that, I was doing so many different things, and I sort of, you know, it was nice to do many different things, and you got to have many different contexts, and you got to do many different types of activities, and when you got tired of one thing, you could go and do another thing. This is not that, right? Um, it is like running a mega sprint as part of a marathon, and I love it, But I think, you know, could I go back and also have that expectation going in? I think it would have made it, I think I would have been better at handling other disappointments or, you know, parts of my life that I couldn't focus as much on.

AI assessment note: “What the intensity of doing and thinking about nothing else would, um, sort of impact”

Q Did they have the idea for Lagor or Lea at the time then?

A The idea for the business was not a product. It was a problem space. It was the fact that AI and legal is going to be a thing. And they had been playing around with the early BERT models that came up from Google, and also a, a version of them called SWE BERT, They were so incredibly bad, frankly. I mean, you couldn't really solve anything back then, and so they had built their own document management system, they had built their own editor, they had built, like, a legal research platform, they had all this code, and when I came in, I sort of said, let's just delete all of it, and let's build on GPT, and let's, like, build a version of it that, at the end of the day, lawyers are benefiting from.

AI assessment note: “The idea for the business was not a product. It was a problem space.”

Q days a week, and the valley has turned up the intensity on work ethic, and everyone killed me for it. I find life very funny. Everyone always recommends the book The Courage to be Disliked, and then when you actually gain that courage, they will shoot you. So my question to you is, do you agree with the nine-nine-six mentality, and how would you reflect on what I just said?

A Yeah. I mean, the short answer is yes, because we did it for a while, and we, we did it with large team buy-in, If you tie it back to our space, our view of it is there's clearly going to be a couple of very valuable businesses in our area, and if we're playing to be number one, we're gonna need to outwork, outdeliver, outcare our competitors, and you can do that through different ways than just pure hours, but I think it's part of that function, of course. Like if you work 10 hours, A day and I work eight. I mean, at some point you're going to have achieved a lot more unless I'm very productive in my hours and I need time to reset. And I think there are people who can be quite different. I think it's different if you're building high complexity systems where you need to think deeply about directionally where you're going, or if you're sitting and answering emails. So there's different work for different places.

AI assessment note: “the short answer is yes, because we did it for a while”

Q days a week, and the valley has turned up the intensity on work ethic, and everyone killed me for it. I find life very funny. Everyone always recommends the book The Courage to be Disliked, and then when you actually gain that courage, they will shoot you. So my question to you is, do you agree with the nine-nine-six mentality, and how would you reflect on what I just said?

A Yeah. I mean, the short answer is yes, because we did it for a while, and we, we did it with large team buy-in, If you tie it back to our space, our view of it is there's clearly going to be a couple of very valuable businesses in our area, and if we're playing to be number one, we're gonna need to outwork, outdeliver, outcare our competitors, and you can do that through different ways than just pure hours, but I think it's part of that function, of course. Like if you work 10 hours, A day and I work eight. I mean, at some point you're going to have achieved a lot more unless I'm very productive in my hours and I need time to reset. And I think there are people who can be quite different. I think it's different if you're building high complexity systems where you need to think deeply about directionally where you're going, or if you're sitting and answering emails. So there's different work for different places.

AI assessment note: “the short answer is yes, because we did it for a while”

Q You said about like partnership and that being very central to how they think and how they choose. You know, I had, um, Matt Fitzpatrick, the founder of, or the CEO of invisible, which is kind of like a McCore or a Turing competitor. And he said that essentially it is impossible to sell into enterprise without an FDE model. Do you agree with that? And are you seeing that?

A We have a very big A team of legal engineers who are ex-practicing lawyers from the top tier best firms, but they're not fully seconded, right? They're forward deployed in the sense that their main job is to make you successful. An example would be, um, a big firm that just went with Legora Widencase. Wide and Case now has the challenge of adopting AI across their entire firm. It's a big firm, such an enormous change management undertaking to equip all the lawyers across all the different practice areas, across all the offices, and across all the skill levels from associate, senior, associate to partner with AI proficiency. And we need to make them successful because if they are not successful on Nagora, um, you know, A year from now, two years from now, it's going to be a really sad conversation. So we invest a ton of upfront manual labor and time and effort in doing the implementation and activation right. So I do think that's necessary for enterprises where you're changing the way they work. If you just think about a process, right? Like let's say you're working in Um, for if we're just staying in legal, if you're working with AI contracting, you basically have a contract lifecycle management system. You just send a document somewhere, generate some red lines, and then you put it back. That's pretty easy. You don't need a forward deployed engineering or legal engineering mod…

AI assessment note: “So I do think that's necessary for enterprises where you're changing the way they work.”

Q So your biggest advice to founders on scaling in the US Without committing large resources would be that you can do a freemium and test it from Europe?

A Yes, for sure. I think, well, I think we, we could, so why can't you? And we're very enterprise. That might be different, right? We did not need to invest a ton in marketing or like B to C content in the US. I could just get on demos and get on a few flights, um, you know, demo the product, run a few pilots. You know, always competitive pilots. And then again, on the partnership level, um, show that we were willing to work with these firms on their ambition level, because it's very high. And the firms that we work with are not treating AI as a check the box exercise. It's not, oh, let's buy, you know, this thing, let's roll it out and we're done. We want to be the firm that Dominates our market because we understand AI and technology better than any other firm.

AI assessment note: “Yes, for sure. I think, well, I think we, we could, so why can't you?”

Q Yeah. In the land grad time, what is the biggest challenge that you face?

A The biggest challenge that We have right now is growing from 30 to 300 and then doubling again in the next two quarters from 300 to 600 and maintaining the ambition, integrity, teamwork, and just like raw grit that got us here when you double the team. I think we just hired two new, um, People in the US and they were really surprised by how late everybody was working. They were like, oh, at the, at the other place I was at, which was a, you know, another, another legal tech provider. Um, everybody left at six. We have dinner in the office at eight. And so when your entire team globally operates at that level and at that pace with that, you know, goal in mind, that's awesome. But I care a lot about maintaining that. I think the other main challenge is there's so many things.

AI assessment note: “The biggest challenge that We have right now is growing from 30 to 300”

Q when you look at the landscape, when you think about kind of that bundling versus unbundling, how you thought about that? When you look at the landscape today, How does that landscape look in three to five years? Is this a winner take? If we, okay, actually a much better way to ask that is, is this an Uber and a Lyft? Or is this a Google Cloud, AWS, Azure?

A The reason why I don't think it's a Uber and Lyft is because in Uber and Lyft, there was no product differentiation. The products were pretty much the same. And it was a strategy differentiation on global versus global. Yeah, but it was very hard to build something different. I mean, Uber for what it's worth, I mean, the product looks the same today, basically, right? With the addition of, you know, bells and whistles, but it's the same fundamental thing. But, um, the difference to our story is that the product differentiation really matters. And there, and the, and the amount of things that you can go and build is so vast. It's like this universe of, uh, legal technology that just has never been built. Because one of my theories is that maybe in legal tech before generative AI, there weren't that many exciting things to build. And it was really hard to scale a good company. And as soon as you got to like single digit million revenue, you would get an acquisition offer, which would be life changing money for the founders, but you know, no unicorn outcomes. So the Product strategy will impact the trajectory of all of the businesses in our vertical tremendously. I think it's totally a winner takes all, like all sauce. I mean, you know this. Number one will grab 90%, and number two to number 10 will share the remaining 10%, and so I think what that means for us is, You gotta run l…

AI assessment note: “I think it's totally a winner takes all”

Q Can you help me? In 24 months, rank the model landscape for me.

A I'll bet based on what I've seen in the last sort of three, six months. I think for our type of work, it will either be cloud or Gemini. That is the top model. And it will be dependent on if context window is a very important factor or not. So far it is not because we've built so much, um, architecture around handling lack of context window that we still prefer the cloud models. Uh, or the anthropic models. Um, and it seems to me like open AI is going down the, you know, let users fine tune models like that type of, um, journey a bit more, which so far I don't have a lot of reason to believe it.

AI assessment note: “it will either be cloud or Gemini. That is the top model.”

Q So your biggest advice to founders on scaling in the US Without committing large resources would be that you can do a freemium and test it from Europe?

A Yes, for sure. I think, well, I think we, we could, so why can't you? And we're very enterprise. That might be different, right? We did not need to invest a ton in marketing or like B to C content in the US. I could just get on demos and get on a few flights, um, you know, demo the product, run a few pilots. You know, always competitive pilots. And then again, on the partnership level, um, show that we were willing to work with these firms on their ambition level, because it's very high. And the firms that we work with are not treating AI as a check the box exercise. It's not, oh, let's buy, you know, this thing, let's roll it out and we're done. We want to be the firm that Dominates our market because we understand AI and technology better than any other firm.

AI assessment note: “Yes, for sure. I think, well, I think we, we could”

Q Yeah. In the land grad time, what is the biggest challenge that you face?

A The biggest challenge that We have right now is growing from 30 to 300 and then doubling again in the next two quarters from 300 to 600 and maintaining the ambition, integrity, teamwork, and just like raw grit that got us here when you double the team. I think we just hired two new, um, People in the US and they were really surprised by how late everybody was working. They were like, oh, at the, at the other place I was at, which was a, you know, another, another legal tech provider. Um, everybody left at six. We have dinner in the office at eight. And so when your entire team globally operates at that level and at that pace with that, you know, goal in mind, that's awesome. But I care a lot about maintaining that. I think the other main challenge is there's so many things.

AI assessment note: “The biggest challenge that We have right now is growing from 30 to 300”

Q What product decision did you make that with the benefit of hindsight was a mistake, and what did you learn?

A The first version of the Lagora product back in summer of twenty-twenty-three, that was completely the wrong direction. We built it centered around a couple of core use cases, and we did not have an agent Or a chat that could operate over those tasks. It was like a click and point use case. Clearly the wrong direction. So after we got accepted into Y Combinator, we deleted all of that code. I think we made some good product decisions. By very early adopting, uh, basically Langchain was too bad at the time, so we built our own agent architecture, and that, we did that very early, which I'm very proud of, and that was like the right direction to continue on.

AI assessment note: “The first version of the Lagora product back in summer of twenty-twenty-three, that was completely the wrong direction.”

Q Do you not need that training to create a pipeline of future partners?

A I don't think you need the training of doing those specific work tasks. I think you need the training on the process and the value you're delivering to the end client and how every piece fits together. But I don't think you need to do the work of learning exactly how everything is phrased, because I think here we're going to rely on AI systems to help us. And I mean, it's, it's a similar parallel to engineering. Like, I don't think you need to understand exactly how every piece of row works if you understand how the system works, because I don't think they're going to pay for it. I think in that future you need to bundle in that no firm is going to pay for somebody to run, go through that education.

AI assessment note: “I don't think you need the training of doing those specific work tasks.”

Q What product decision did you make that with the benefit of hindsight was a mistake, and what did you learn?

A The first version of the Lagora product back in summer of twenty-twenty-three, that was completely the wrong direction. We built it centered around a couple of core use cases, and we did not have an agent Or a chat that could operate over those tasks. It was like a click and point use case. Clearly the wrong direction. So after we got accepted into Y Combinator, we deleted all of that code. I think we made some good product decisions. By very early adopting, uh, basically Langchain was too bad at the time, so we built our own agent architecture, and that, we did that very early, which I'm very proud of, and that was like the right direction to continue on.

AI assessment note: “The first version of the Lagora product back in summer of twenty-twenty-three, that was completely the wrong direction.”

Q What do you know now that you wish you'd known at the start of Leia?

A What the intensity of doing and thinking about nothing else would, um, sort of impact your own psyche and personality. For the last two and a half years, I've basically done nothing else than to think about Leia or Legora or the business, and You know, when you're in, when you're in college, or prior to that, I was doing so many different things, and I sort of, you know, it was nice to do many different things, and you got to have many different contexts, and you got to do many different types of activities, and when you got tired of one thing, you could go and do another thing. This is not that, right? Um, it is like running a mega sprint as part of a marathon, and I love it, But I think, you know, could I go back and also have that expectation going in? I think it would have made it, I think I would have been better at handling other disappointments or, you know, parts of my life that I couldn't focus as much on.

AI assessment note: “What the intensity of doing and thinking about nothing else would”

Q Do you light the tinder, so to speak, and fuel the fire?

A Oh yes. Of course. I think I'm very good at it actually. Um, and competition can be played at a macro level where you think us versus them, but you can also do it at a lower level, which is our marketing team wants to be that marketing team, or, uh, our engineers want to build a faster document upload time than that other team. So you compete on all these micro levels and you celebrate them like crazy, right? I think what I've learned this year, I actually used to be quite bad at celebrating. Um, I remember when I was in, in business school, my, my dream job was to go to McKinsey because I thought that's where, that's where all the amazing people went. Um, I found out maybe that that was not the case, but when I, when I got the call and I got the job, I was in the, Um, in the grocery store, and I celebrated by buying a bag of peanuts.

AI assessment note: “Oh yes. Of course. I think I'm very good at it actually.”

Q Do you not need that training to create a pipeline of future partners?

A I don't think you need the training of doing those specific work tasks. I think you need the training on the process and the value you're delivering to the end client and how every piece fits together. But I don't think you need to do the work of learning exactly how everything is phrased, because I think here we're going to rely on AI systems to help us. And I mean, it's, it's a similar parallel to engineering. Like, I don't think you need to understand exactly how every piece of row works if you understand how the system works, because I don't think they're going to pay for it. I think in that future you need to bundle in that no firm is going to pay for somebody to run, go through that education.

AI assessment note: “I don't think you need the training of doing those specific work tasks.”

Q Okay. So when we look at the numbers, like bluntly, the retention for Harvey is, it's 98% logo retention, 170 eight percent net revenue retention. Do you have as good numbers?

A So for both those numbers, yes. But on NRR, I don't think that's a fair number for me to comment on because there's so much, so much of our growth is not about renewing contracts from 20 24. Right? Like we did a, in a single day in 2025 in December, we added seven million of ARR, like one day in 24 hours. And that was more than what we did in 2023 and 20 24 combined. And so, NRR and logo retention, you know, it's up to 2026 to determine where those, you know, real numbers will be. And I think for, for what it's worth, the ability of these products to go quite broad Will be very interesting because to some extent there's initial use cases that you can solve with AI that we target. And then the more time we spend with our clients, the more problems and opportunities we see. And so what's happening to the product is they're growing quite a lot. And so what I think will happen is that this will be like a suite, like a platform kind of play that just becomes more and more of the central System where they do their work.

AI assessment note: “So for both those numbers, yes.”

Q Um, I, I'm going to ask you, have they ripped your product?

A Well, I think we take a lot of pride in developing our product as fast and as well as we can. And I think there's two main parts to our product development. One of them is improving the parts that we have, and the other one is making new qualified bets. And I think we have had a history of making bold and correct bets. I think there's many legal tech products that on the surface looks pretty similar. There's even another product where they ripped our name and it's, it's our tabular review. It's just called tabular review in their product, which is totally fine. Um, but what happens when you then go into these competitive pilots and the user starts to kind of rip them apart, that's where you see that, you know, One product is a Rolls-Royce and the other product, or maybe a Volvo, and the other product is, is, is maybe a cheaper version.

AI assessment note: “There's even another product where they ripped our name and it's, it's our tabular review.”

Q Um, I, I'm going to ask you, have they ripped your product?

A Well, I think we take a lot of pride in developing our product as fast and as well as we can. And I think there's two main parts to our product development. One of them is improving the parts that we have, and the other one is making new qualified bets. And I think we have had a history of making bold and correct bets. I think there's many legal tech products that on the surface looks pretty similar. There's even another product where they ripped our name and it's, it's our tabular review. It's just called tabular review in their product, which is totally fine. Um, but what happens when you then go into these competitive pilots and the user starts to kind of rip them apart, that's where you see that, you know, One product is a Rolls-Royce and the other product, or maybe a Volvo, and the other product is, is, is maybe a cheaper version.

AI assessment note: “There's even another product where they ripped our name”

Q Okay. So when we look at the numbers, like bluntly, the retention for Harvey is, it's 98% logo retention, 170 eight percent net revenue retention. Do you have as good numbers?

A So for both those numbers, yes. But on NRR, I don't think that's a fair number for me to comment on because there's so much, so much of our growth is not about renewing contracts from 20 24. Right? Like we did a, in a single day in 2025 in December, we added seven million of ARR, like one day in 24 hours. And that was more than what we did in 2023 and 20 24 combined. And so, NRR and logo retention, you know, it's up to 2026 to determine where those, you know, real numbers will be. And I think for, for what it's worth, the ability of these products to go quite broad Will be very interesting because to some extent there's initial use cases that you can solve with AI that we target. And then the more time we spend with our clients, the more problems and opportunities we see. And so what's happening to the product is they're growing quite a lot. And so what I think will happen is that this will be like a suite, like a platform kind of play that just becomes more and more of the central System where they do their work.

AI assessment note: “So for both those numbers, yes.”

Q Can you help me? In 24 months, rank the model landscape for me.

A I'll bet based on what I've seen in the last sort of three, six months. I think for our type of work, it will either be cloud or Gemini. That is the top model. And it will be dependent on if context window is a very important factor or not. So far it is not because we've built so much, um, architecture around handling lack of context window that we still prefer the cloud models. Uh, or the anthropic models. Um, and it seems to me like open AI is going down the, you know, let users fine tune models like that type of, um, journey a bit more, which so far I don't have a lot of reason to believe it.

AI assessment note: “I think for our type of work, it will either be cloud or Gemini.”

Q Do you light the tinder, so to speak, and fuel the fire?

A Oh yes. Of course. I think I'm very good at it actually. Um, and competition can be played at a macro level where you think us versus them, but you can also do it at a lower level, which is our marketing team wants to be that marketing team, or, uh, our engineers want to build a faster document upload time than that other team. So you compete on all these micro levels and you celebrate them like crazy, right? I think what I've learned this year, I actually used to be quite bad at celebrating. Um, I remember when I was in, in business school, my, my dream job was to go to McKinsey because I thought that's where, that's where all the amazing people went. Um, I found out maybe that that was not the case, but when I, when I got the call and I got the job, I was in the, Um, in the grocery store, and I celebrated by buying a bag of peanuts.

AI assessment note: “Oh yes. Of course. I think I'm very good at it actually.”