'The Hardest Thing I've Ever Done': Noah Waisberg on Why Selling a Business Hurts, and How Zuva Aims to Fix It

0:02Noah Waisberg: Welcome to Law Next. Five years ago, Noah Waisberg sold Kira Systems, the pioneering AI contract analysis company that he co-founded in 2011, to Latera. Even though the deal itself went well, the experience of going through the sale was hard, Noah now says, and it inspired him to turn that experience into his next business. As part of that sale, he kept a copy of the underlying Kira technology and about 30 of the company's roughly 180 employees, and he used them to launch his current business, ZUVA. ZUVA started as an API-only company, but it has since turned its focus to the sell side of M &A.

0:41Noah Waisberg: After interviewing some 200 dealmakers about the most painful parts of selling a business, ZUVA built software to help sellers manage the flood of bitter questions. It also launched a tech-enabled diligence service that pairs Kira's AI with experienced former Big Law associates. Zuva says the service delivers work comparable to that of Big Law at a fraction of the cost. Noah was last on Law Next in 2021, around the time his book AI for Lawyers came out. In today's episode, he explains Zuva's new direction and what the recent expiration of his post-sale restrictions means for the company. We recorded this conversation on the day Harvey announced its latest mega funding round.

1:27Noah Waisberg: So we also talk about the prospects for Harvey and Lagora against the Frontier Labs and why Noah told me at Iltecon that we've hit peak legal tech. He also covers the limitations of today's AI that he thinks the market underestimates and why the billable hour remains the biggest obstacle to efficiency in law. This is Bobby Ambrose, you're listening to Law Next, the podcast that features the innovators and entrepreneurs who are driving what's next in law. And before we get to any of that, please take a moment to hear from the sponsors who so generously support Law Next.

2:07Noah Waisberg: This episode is brought to you by BriefPoint, the discovery governance and automation platform used by more than a thousand law firms. BriefPoint just launched Discovery Playbooks. Here's the idea. You respond to Discovery like you always do. Save it to your playbook in one click. And the next time a matching request shows up on any case, BriefPoint reapplies your objections and response strategy automatically. No data entry, no setup. Watch the 60-second demo at BriefPoint.ai. Streamline your law firm operations with Practice Panther, the all-in-one legal case management software. Trusted by tens of thousands of lawyers, Practice Panther's easy-to-use features, such as customizable workflows, automated billing, and seamless client communications, help you accomplish more in less time.

3:02Noah Waisberg: Making the switch to Practice Panther is easy, with dedicated data migration, live training, and support. get your free trial today at practicepanther.com. Now let's get to today's conversation. Noah, welcome back to Law Next. Good to see you, Bob. Thanks for having me again. I say welcome back because you were on the show in 2021 on the occasion of right around the time, I think it was right around the time, your book came out, which I have people watching on video can see I've got Noah's books right here in front of me. I was looking, I'm kind of disappointed. I got my little autograph in the book.

3:40Noah Waisberg: I don't know how that happened. You've got a few of them up there too, I'm sure. I got boxes sitting at my feet, yes. But it's kind of funny because you literally wrote the book, AI for Lawyers, and lots changed probably since then. Have you ever thought about redoing it or maybe not? Do you think a lot's changed? Does it need an update? A lot has changed and a lot hasn't changed. the central thrust of AI Lawyers, the book, is that an hourly billing lawyer can make more money being more efficient. Like that is the core intellectual idea of the book. And it's possible that I am one of the best people in the world at explaining that concept, even though I haven't really done it very much for the last five years as since selling Kira.

4:33But that is the core argument of AI for lawyers. And I don't think that's changed. And I think a lot more people are spending a lot more time looking at that right now. There are some case studies in the book, including by some terrific people talking about specific areas where AI is being applied in law. So we've got like a section by Michael Mills on expert systems. We have Jake and Pablo and Laura, I think, write a section on AI and legal research. There's a Jack Newton subsection, there's a Jason Barnwell subsection. Not all of those, some of the Blue Jay people do a section. Not all of those subsections, I think, are as au courant.

5:23Like if you were going to write a state of the world in AI today, I think it'd be different. But to me, when we were writing the book, we were intentionally, we knew that AI was going to change. And so there's sort of like a part one that talks about the intellectual case around AI for lawyers. And I think that if anything is more powerful today than it once was. There's a second section of the book that's about case studies. And I think if we were to update it, and we may, that is something that would have to change. And then there's some other sections of it that I think are also more lasting about how AI is played out.

6:01So I think there's things we'd have to change, but actually, I don't think it's that much that we would. And it's something that we've talked about. I've just had my hands a lot more full than I've expected.

6:15Noah Waisberg: Well, that's a good thing, I guess. I'm sure it is. Yeah, mostly, but probably 80 % good. people watching this, I might encourage them to go back and listen to that earlier podcast. You and your co-founder, you and your Kira co-founder and Zuva co-founder on that. But that was when you were still at Kira and on the show and talked a lot about the early history and the founding and the development of the company. But I was thinking actually in thinking about today how it's really interesting, and I'm sure you're well aware of this, but how it's sort of like each stage of your career has provided the fuel for the next stage of your career.

7:03Noah Waisberg: I mean, you were a diligence lawyer, a practicing lawyer, and then you saw a need and developed Kira, which is a software for helping buyers perform diligence reviews. And then you sold Kira to Latera and you went through the process of being a founder selling your business and then you started ZUVA and you've been kind of focusing more on the sell side, I think, since going through that experience and starting this company. So I don't know, does that make sense? Do you kind of see your career? I probably should just, I don't know, maybe I'm not that creative. In fact, actually, I probably am thinking about businesses all the time, right?

7:56Like, I'm definitely someone who sees the world and is thinking about businesses to better issues that I see in the world. But I think, I guess the ones that I feel like I have the most to bring to the table on are ones where I've had the experience myself, but maybe not just had the experience myself, like, really, really gone through the experience. And I think this specific niche of M &A has been something that I've really been doing since 2006, 2007, like when I was a lawyer.

8:30Noah Waisberg: 2011 is when you started working on Kira, right? Yeah, totally. But I started as a corporate associate in 2006. And when I was doing that work, that was mostly private equity and M &A work. and um and in that you just realize that m &a is this fantastic area to work in that there's really smart people in it really interesting deals interesting projects and lots of demand and frankly also lots of spend right like a lot of desire to get it right like you think about how good times large law firms are having right now. And that's like very heavily M &A and PE driven, right? Like that same thing I was doing back in 2006, 2007, 2008 is still the thing that's powered these firms to greater and greater heights.

9:29And so it's kind of an exciting place to work in. And I think an area where there's still, having been through the experience of Selleagera, a lot of room for improvement. And it's really like, I think many people would describe selling a business as one of the hardest professional things that they've ever been through. And that's the case as much today as it was a decade ago, 20 years ago. And that's all I really have experience with, but presumably 20 years before that too.

10:00Noah Waisberg: What makes it hard? Is it the legal and economic side of it or is it the personal side of it? uh there's a lot of things that make it hard like you you work for like with kira we worked for about a decade before we sold it right like 10 and a half years before we sold it right and it's like uh you run a business like you know that if you're running a business it's not something where even if it wasn't your business like if you're a manager of business like you spend a vast amount of your life thinking about work and it's like something you're so bound up in like how do I do this better and the sort of selling of it isn't something you get to do that often and it really can be big difference how will you do it so I think one part is this just transition of phases I think it's hard the other part is just as a practical thing you know we all have day jobs like your day job takes you 40 hours a week 50 hours a week 60 hours a week um selling a business is something that can take an additional 40 to 60 to 80 hours a week um but you can't stop doing your day job either right so and frequently if you're in a bigger enterprise you can't even tell the people most of the people that you're working with about what you're doing right like When we were at Kira, we were about 180 people when we sold and probably 15 people inside the business for most of the deal who knew what was going on.

11:36Like initially, maybe only five. And then as things got sort of farther along, maybe we expanded the tent up to about 15. but you go from that means there's you know 165 people 170 whatever who don't know what's going on inside uh the thing who expect you to get things done and you can't do it and in the meantime you have to do all sorts of administrative things yourself right to get the deal over the hump and you don't have the support that you need and even like the people who really know how to do this stuff don't know about the deal and can't know about the deal and so it makes it pretty hard uh because of that so actually like as we as we got going on the thing that we're most focused on right now in zuva which is selling a business what we did is we went out and we spoke to 200 ish m &a people about uh pain in selling a business so specifically selling kira After selling Kira, we kind of, we're like, we're selling an API only version of Kira.

12:49It's going okay. We still have customers on that. Still think it's a useful product.

12:53Noah Waisberg: Because you launched Kira Suva almost at the same simultaneously, I think. Five years ago. Like it just, it literally, as part of selling Kira, we got to keep a copy of the underlying Kira AI. And we got to keep about 30 of our 180-ish people. And that's what became Zuva. And we initially, we had some customers for Kira who were using Kira as a backend to their system. So they didn't care about the Kira UI. They just cared about the underlying Kira AI, which was great at finding data in contracts. And so we launched an API product where those customers could just interact with their underlying technology and not have to use the Kira interface, not have to go through contracting with Kira, security reviews, anything like that.

13:44Just other tech vendors or companies could go and build that AI into their own system, which turned out to be, you know, a model that had legs in the sense that that's kind of what everybody's doing now with open AI and entropic type products, right? Unfortunately, the Kira technology was and is great, but, you know, not as fully featured as what you get out of some of the Frontier Labs. But we were selling that API product and we realized it was going to be okay as a business, but just never, never great. And so we kind of stepped back and we were like, what are, what do we really care about?

14:29What do we think we can really do well? and M &A sort of hit the spot. And specifically selling a business kind of spoke to us because it was something that we thought about a lot at Kira, but we just weren't in a position to really sell to. And at Zuba, we had had this unpleasant, not like anything went wrong, like the deal was fine, Latera and HG were great to work with through the deal, good lawyers, good professionals involved in the deal. it went okay but it was still just really hard on me on our cfo on a few other people in the business we're like if it's this painful for the senior executives of the business like us on a deal that went fine where we've been in the business the whole time we wondered if other people had had the same problem so we went out and we asked and we spoke to 200 people and m &a people, so executives at businesses that it sold, so CEOs, CFOs, GCs mostly, M &A lawyers, investment bankers, and private equity people, as well as other sort of assorted M &A people like reps and warranties insurers or quality of earnings providers.

15:43And we asked them two basic questions. Number one, is it a pain to sell a business? And number two, if it's a pain to sell a business, what are the most painful parts? And happily, we heard very consistent answers across people. So is it a pain to sell a business? It turns out we were getting quotes like, it's the hardest thing I've ever done professionally. Like, literally, we even saw tears on one of the interviews that we did. And so that, as much as you don't like to see people tearing up about your problem is something you do feel like seeing if you're trying to build a new business. So that was question number one.

16:25And then question number two is, what are the most painful parts? And there again, we saw very consistent answers, that there were sort of four areas where we heard repeated complaints. And we thought that because we were hearing such pain and such consistent pain, that this could be a really interesting place to work in. And so that's kind of what we've been up to for the last while. Right.

16:51Noah Waisberg: So you have evolved from the API, the back end of Kira's AI to what? Exactly? What do you now put on? What are the offerings? So we still offer the API. If anybody is looking to pull data out of contracts, our belief is that it's a really nice way to do that. And it fits really well, actually, with gen ai systems like how we use our own technology ourselves these days is we use that underlying sort of cure tech as we've improved it over the years to find relevant parts of contracts and then we put it into gen ai to do a second level filter on top of that to interpret the meaning of what what it found and maybe clean up false positives and just just give you one answer of say can do you need consent uh on change of control or something like that so we still have the api but our focus these days is on two specific or let me tell you about the four pain points that we heard and then i can tell you about what we're doing about it okay okay four pain points were pain point number one is companies need to find a selling company needs to find documents that acquirers would care about.

18:12Okay. And they need to put those in a virtual data room and finding those documents and then putting them in the virtual data room and organizing them is really painful. Finding the documents for the company, if they're assisted by investment bankers, it tends to be junior investment bankers who have to do the most of the organizing. And to be honest, like most people are less concerned about the quality of life of those junior investment bankers. So you wouldn't focus too, too hard on that specific part of the problem. Second problem where we heard a lot, lot, lot of pain was on managing bidder questions.

18:53That each bidder for your business, if it's a decent size sort of cura, like someone's paying like nine-ish figures for it, someone's going to ask 200 to maybe even 2 ,000 questions about your business, each bidder. And you can have, it would be pretty, it would not be weird to have three bidders, right? Or five bidders, each asking 200 to 2000 questions. A lot of those questions, by the way, are going to be duplicative. And some of them are going to be easy enough that they probably should have never been asked in the first place. Like the asker should have just gone in the data room and figured it out for themselves, but they don't.

19:34And because they're going to pay you like nine figures, you suck it up and you answer those questions. The third thing we heard pain around was disclosure schedules. So disclosure schedules are something when you sign a merger agreement, a purchase agreement, there's a part of the document that's pended to it that says details about your business and problems with your business, right? So something like the merger agreement will say, you know i promise that my business has no lawsuits against pending or threatened lawsuits against it apart from those mentioned in schedule you know 7.9 b romanette 3 or something like that and then schedule 7.9 b romanette 3 has to list all your pending and threatened lawsuits that's actually not that big of a deal to do okay like uh or because ideally you have somewhere in the general counsel's mind or a database it's a big enough company a sense of if anyone's suing you and you can write that down right and so that doesn't take that much time another one of the schedules might say none of our employees have a salary compensation over three hundred thousand apart from as listed in schedule.

20:56And same thing, not that hard to do. You get an output from the HR system. You figure out that the employees have a salary above 300 ,000 or comp above 300 ,000 and you put them on the list. But some of those questions, about 20 % of the questions in a disclosure schedule can be really painful to get the information for. A lot of those happen to be contract questions. And what we heard repeatedly was that disclosure schedules end up being the long pole in the tent where the company are typically on the hook for doing the disclosure schedules themselves comes to the week of signing and it turns out that the company has not finished the disclosure schedules at which point a bunch of things happen um number one the deal maybe gets delayed number two all-nighters get pulled number three, their outside law firm goes in and does a ton of work to get the disclosure schedules finished.

21:54But actually, it turns out that the outside lawyers, even at really fancy firms, aren't that happy with the process because it blows the budget and the clients can be sort of low to pay for it afterwards. Like the realization rate on that work is low. So that's disclosure schedules, a lot of pain there. And then a fourth thing we heard was around pre diligence, where I would suggest if you're going to sell a business, that you should do diligence on yourself before the potential buyers, like those bidders who are going to go into your virtual data room, take a look at your business so that if there is anything bad there, you know about it before the people who are looking at your business know about it.

22:41Just as like if I was going to sell a house, I think it would be to your best interest to know if there was asbestos somewhere, right? Right, right. You can wait for the bidders to discover it. But if you do, it has a bunch of adverse consequences, right? Like it can cause certainly a lack of trust, like either trust in you as a seller, like maybe they think you're a sketchy used car salesperson, or maybe they just think you don't even know what's going on inside your business. Neither of those are great. and that can lead to a lot of problems right like it can cause uh just more questions like tougher questions from the bidders it can lead to retrades on price and it can even lead to delays or deals going away and so i would suggest if you're going to sell a business which is a massive amount of work that you should know that stuff in advance and in fact not only would noah like ex-M &A associates suggest that.

23:39But if you talk to partners at high-end law firms, they will also say that you should do this work. The interesting thing we found though is you ask a partner at a high-end law firm, like, do you tell your clients to do this? They're like, oh, totally tell our clients to do this. And then you ask question number two, which if you're me, you're asking this question and you say, do your clients do it? And they're like, well, not really, or almost never. And it's like, why don't they do it? And it turns out they don't do it because you got to do the work before you actually know that you have bidders.

24:14Like if you were going to sell your business, you have to do it before you know that someone is going to buy it. And it's a lot of money. And a lot of people are unwilling to sort of do that spend before they know that there's a legitimate buyer on the other side. and in fact they actually have to do the work anyways at some point to populate the disclosure schedules that we just talked about but they're sort of unwilling to do the work in advance when it could actually make a difference for them on selling and they're unwilling to do it because of cost so with those four problems in mind what we've been doing is two things number one we've been building software that helps people manage bitter questions so you can dump spreadsheets or word documents or whatever in software will create a database of those not too complicated it'll tell you that question 42 from bitter a is really similar to question 56 from bitter b and here's how you answered question 42 which can be time and annoyance saving and also enable you to generate more consistent results.

25:24And then we're actually getting into the point where we're starting to try to help people answer the questions themselves, where the questions are easily answerable. And I think one of the things that I was like...

25:36Noah Waisberg: Is it also identify all those duplicative questions that are going to come from multiple bidders? Yeah. So it'll identify the duplicative ones and it'll actually go one step further, which is say, like, here's how you answered this last time. Would you like to just reuse that answer? And then it'll also try to answer, help you answer questions yourself where it thinks it can. So that's one of the things that we're doing. And that one, we would sell it to companies who are selling themselves. Generally, we'd reach them through investment bankers. So we've gotten to know quite a lot of investment bankers around this.

26:14And they would suggest to a client who's selling that this could be a helpful thing for both the bank and the selling company to use this software. So that's number one. And then the second thing that we've been doing is we realized that as part of the sale of Kira, we got to keep a copy of the underlying Kira IP, which is still my sense when I look at those sort of stats, the sort of premier software for doing M &A due diligence contract review. which by the way is mostly done in North America on the buy side. So we have a copy of the technology that the buy side law firm is reasonably likely to be using on a diligence project, number one.

26:56Number two, over the course of building Kira, we had just a great experience hiring mid-level to senior associates who were from high-end big law firms who were looking for a change of pace. and generally not always, but pretty often people with sort of family responsibilities as well, like really took parenting seriously in a bunch of cases and didn't feel that they could work in big law because of that. And we were able to hire really excellent people who just were available to us because of their family kind of commitments. And we realized that if you took the software and you paired it with typically if a big law firm was doing this they'd put like a first year on doing the diligence work but if you paired it with ai with more experienced lawyers that you could actually it turns out generate diligence reviews that are around the same work product quality for about 80 % less cost than what you get at a big law firm.

28:09And what we found is that that's pretty appealing, including actually, interestingly, to lawyers who've been passing us on to clients, right? Saying that this is something where you client might benefit from this. So using lawyers as not a customer of ours, but more of a referral channel. yeah so we would do uh disclosure schedule prep we would do pre-diligence uh reviews and then we've actually been getting pulled lately into doing buy side uh sort of reviews as well just in situations especially auction situations where you have multiple bidders and the bidders aren't sure if they're going to win the auction so they don't really feel like spending that much money on doing diligence, but if they do win, they feel like having done diligence.

29:05And so in that type of situation, again, where they'd like to be able to do the work, but do the work for a little bit cheaper, we've been finding, we've been able to pick up review projects there. Does that make sense? So basically two services, both around sell side, a bit of, I said, M &A, one is software that helps you answer the bitter questions. And the other one is a tech enabled service around helping companies deal with diligence. So both really around diligence, but different flavors of that. Yeah.

29:41Noah Waisberg: No, yeah, that totally makes sense. I know that one of the reasons that you, after you sold Kira and started Zuba, one of the reasons you kind of focused on that API product was that you had restrictions in your deal on developing UI-based products, right, that kept you from doing that. And that's due to expire this year, as I understand it. It says so on your website. I'm not giving away any secrets here. Yeah, no, so in fact, yeah, we had some restrictions. I think they're pretty typical to have in the type of situation that we were in. The restrictions have expired, and we're now like, you know, free and available.

30:28But that said, I don't know that our, you know, I don't know that there are ways that could change things for us. It certainly opens up possibilities and just, you know, means, you know, in the past, anytime we talked to one of our friends at a law firm, we'd have to say, hey, we can't sell to you, but, you know, whatever, or something like that. or we can't, sorry, I can't give you advice on this topic, that kind of thing. And it makes things a lot easier for me to not have to sort of go through that anymore. But I don't know that it necessarily changes what we're going to be doing in terms of business.

31:09Like certainly we could sell against Kira now if we felt like it, but I don't have some, you know, hateful chip on my shoulder. Like I got a debt.

31:21Noah Waisberg: But I got to get out. You weren't just waiting for this moment to... Yeah, no, no. I'm not like Parker Conrad. You weren't just waiting for this moment to pull a new product out of your hat. You know, things I've not been wronged or anything like that. But yeah, the restrictions are now gone. And I am happy to be liberated from that, even though as a practical thing, my focus is on the same thing today as it was a month ago, which is just trying to sort of build our company the best way that we can. And we're really excited about the sell side and going direct to clients. One of the big frustrations I always felt at Kira was that we sort of like by selling to the lawyers in the process, you're really leaving a lot of value on the table.

32:10Like if I think about Kira, everybody would have said Kira helped you review a contract in 50 % less time. Like if we go into AI for Lawyers, the book, there's a case study in there where Meredith Williams-Range was now at Gibson Dunn, but at the time was the chief client value officer of Sherman, said that talked about how when they did a test at Sherman, they found that people who, a head-to-head test of people reviewing contracts using Kira versus reviewing them the sort of manual way, that they found that the people who used Kira reviewed the contracts in 50 % less time and they were more accurate doing it.

32:54And I think that would have been that kind of industry consensus around when we sold the Kira business that more or less anybody could save 50 % of the time using Kira. The interesting thing, so you think about 50 % of the time, right? And if I think about Latham was a Kira customer and Latham charges like a thousand bucks an hour more or less right now so if you save 50 percent of a Latham hour then that's like five hundred dollars right like assuming it's an hour to review a contract which I think is a pretty good assumption generally it's about three minutes a page for a person to manually review a contract and I would say a typical contract is 15 to 20 pages so you know 45 minutes to an hour i think is a pretty typical range but just to make math simple let's go with an hour right so you save 500 bucks a contract if you're using kira and if you're doing 100 contracts that's a 50 000 savings you're doing 200 contracts that's a hundred thousand dollar savings to someone um but to someone is not lathom like lathom would not come close like anyone at a firm would be flabbergasted at the idea of spending you know even 50 bucks a contract for that review right like latham would not pay kira or harvey or anyone 50 bucks a contract to review a contract for them right but the actual economic value to the client is like 500 bucks of using Kira, even putting aside the accuracy thing.

34:37And so the interesting problem to me for the last bunch of years has been like, how do you get closer to being able to get paid for some of the value that you're bringing to the world? And I think one of the interesting things is not necessarily by doing it from taking money out of the firm's pocket, which I just don't think helps you because the law firms are such an important part of the ecosystem. But what I'm happy about with Zuba is just trying to do it in a way that actually works well with other participants in the ecosystem, like investment bankers and M &A lawyers.

35:23Noah Waisberg: So far in our conversation, Noah has explained how the pain of selling Kira shaped what ZUVA builds today, from software for managing bidder questions to an AI-enabled diligence service aimed at sellers. When we come back, we'll step back to look at the broader legal AI market, what made Kira's AI hard to beat, whether Harvey and LeGorak can hold their own against the Frontier Labs, and why NOAA thinks we've reached peak legal tech. But first, please take this opportunity to hear about the sponsors whose generosity supports this podcast.

36:03Noah Waisberg: Practice Panther is the leading practice management software designed to help you analyze, automate, and grow your firm. Intuitive design and easy-to-use features like customized workflows, automated billing, and client portals will help you streamline your practice. Plus, dedicated data migration and live training will help you get up and running quickly. Get your free trial today at practicepanther.com and see why Practice Panther is the smart choice for lawyers. Again, that's practicepanther.com.

36:41Noah Waisberg: Welcome back to Law Next, where my guest guest is Noah Waysberg, the co-founder of ZUVA and earlier of Kira Systems. Before the break, Noah walked us through ZUVA's focus on the sell side of M &A. I want to go back to Kira's early days when AI-powered contract review was something almost no one else was doing, then bring it back to the current state of legal AI. Let's get back to that conversation.

37:11Noah Waisberg: some people who are listening to this conversation right now may not realize that kira when it law when you launched kira when you developed kira was unlike anything on the market there there this this was a a category that hadn't been invented yet i mean now it's it's kind of table stakes, right? That people are going to use diligence review and contract review software in deals. I don't even know if we talked about AI back then. I mean, we must have, I guess. I don't know. Did you even describe it as an AI product back then? I don't remember exactly. In the earliest days, we would have called it machine learning contract analysis.

37:53But up here also that sort of this thing is my 2015, Robbie the Robot Learns to Read, which as far as I know is the world's first children's book on AI, right? So when we're talking about, we have been talking about AI for a fair while. I think most people that I've spoken about it would say that Cura was the leading legal AI business when we sold it in 2021. Yeah, no, I think

38:25Noah Waisberg: that would have been a fair statement certainly from my observations that was that was certainly the sense of the market um one of the really just taking off on what you just said bob though one of the really interesting things about ai as we built it to kira versus the ai today is like a lot of the ai today like the actual ai is the same ai right like you're building claude you're building gemini you're building gpt Maybe you're building something else like Quen or Grok or like an open source model into your offering. But fundamentally, you're able to use the same tools that everybody else has.

39:12And I think some of the larger companies like Harvey keeps putting out really interesting research about trying to like get beyond just using those same underlying pieces of technology. and building different harnesses and the like. And I assume Lagora does the same. I just happen to have seen for everyone earlier today. But the sort of fundamentally, like the AI is not going to be that different using one tool versus another tool today. Whereas for us, we had put a lot of work into developing our AI. And it just meant that like we had like a decade probably where our results were just better than anyone else's and that's a really nice situation to be in when you're selling things it gave us much more of a durable competitive edge like it was kind of something we like some of our competitors i think they're really nice and they had in some cases and they had really good features sometimes that we didn't have but ultimately like what we kept hearing and this is me like maybe uh again have a different view but what we kept hearing was that uh our ai like the underlying accuracy of our results themselves were just better and that was really hard for other competitors to kind of match up with especially as we started getting more money and were able to invest even more in bettering our ai it just became something where we had an edge.

40:47And now you see a situation where even like the foundation models, model developers don't really have a durable competitive advantage, right? Like today we're all talking about Astra, but in two weeks, you know, it'll be whatever comes after Fable and so on and so forth. And Google will come up with something.

41:05Noah Waisberg: So what does that mean for some of the legal domain specific AI providers out there? I mean, And we're talking, recording this on the day that Harvey just announced another$550 million raise to, you know, what was its new$15 billion valuation? I have to say, by the way, just as like a separate aside here, like what Harvey and Lugar have done in terms of growing revenue is like other things besides growing revenue, but just growing revenue and growing customers is exceptionally impressive. Like I'm so awed by what both of them have done there. Yeah, what does all this mean for them? We'll see. I don't know.

41:49I think there's so many ways it could play out right now. Like I think there's sort of the pitch that at some point OpenAI and Anthropic are going to fully focus on legal. And when they do, it's, they're going to be able to, however much, like$550 million is sort of objectively a lot of money, but it's actually not that much money to an OpenAI or an Anthropic or a Microsoft, like those places can spend that in a second, right? and they can, if they really decide to focus, if one of the XL more broadly focused companies decides to focus on the space, that they'll be able to build great things. That would be kind of like the bear case against Alagora or Harvey.

42:41But what I've experienced in using Claude, in using GPT, and I use both like a ton daily basis, is that there's a lot of like engineering challenges there, things that they haven't built out that I think are relatively essential for doing large-scale legal work, which I'm not really doing, but I kind of sometimes I'm on the fringes of just for stuff we have to do ourselves or the like. And I can see how however much resources OpenAnthorpic have, they haven't been able to put in the effort to build those specific things. And they haven't been able to put in that effort because they have lower hanging fruit.

43:27Like there are things that they can do that they think are better. And I think for Harvey and Legora, as long as they can kind of, as long that situation holds, I think there's other ways that they can do well as well. But I think there is a situation where I can see them continuing to be really valuable companies. You know, should any company be worth what it's worth? Like, it's so hard to tell, you know, what any company should be worth at the time that they kind of realize that value. But I think there's sort of like a story where they could be really threatened. And I think there's another story where they could be really valuable, but you've got sophisticated customers paying Harvey.

44:08Like the impressive one to me is not the$15.5 billion valuation. It's the$400 million of revenue on Legora putting up really impressive revenue numbers as well. And I think, you know, these customers are pretty sophisticated if they thought they could get the same thing at a clod they would get the thing at a clod like they're not doing it to get a press release these days like maybe in the early days it was worth something to have a harvey press release but you'd actually probably get more press right now if you're a large law firm and you said we're just going straight clod and we're not doing harvey and we're not doing lagora like that would probably be a more pressy announcement so i think to the extent sophisticated customers are paying for these systems, there's probably a reason that they feel that it's worth it.

44:56Noah Waisberg: Although one of the theories I heard thrown around at Iltacon, you and I were both at Iltacon, was that some of these larger firms with the budgets to be able to do it are kind of hedging their bets. They're not going with one over another because they're not sure which one, if any, is going to win or lose. And they want to be positioned to keep going forward, no matter which one that is. I wonder how much that's the case. Like, so on the one hand, that could be a stable situation, right? Like, certainly, I don't know how much it happens to that you'd have a better idea than me. But back when I was at while got y 'all uh certainly we would have paid for both lexus and westlaw um and probably blue or like whatever else right and it was pretty immaterial in the context of the overall business um some of those charges got so i could imagine a situation where people pay for both i don't i wonder how much that's the case though i think that if i were a firm uh i think firms have gotten more or most firms are pretty sophisticated about this i think they realize that the biggest problem is just getting your attorneys to use any piece of technology and uh i think if you kind of confuse like which is one reason why i think changing vendors is such a thing that they're very careful about uh because they know how much effort it is to get attorneys to use any piece of technology.

46:33And if you make any changes there, there are a lot of, there's some very tech sophisticated attorneys, but I think a lot of people aren't paying that much attention to IT and would just like it to be there and work. And if you start changing around applications, I think attorneys would be not too happy. And I suspect that there's not a ton of firms that are seriously pushing both Lagora and Harvey. Like there may be ones that have both. there may be ones that have basically bought in on say harvey or bought in on lagora and they may have some licenses for the other one but i think there's a big difference between having like a thousand licenses of harvey and a thousand licenses of lagora versus having like a thousand licenses of harvey and having like 20 licenses of lagora and i suspect that the sort of double using firms are more in the second category than the first.

47:28And they may be keeping their options open. Like if they thought that one or the other was significantly better for them, they might make a wholesale shift. But I think it's not that easy for them to shift over, even putting aside any of the like data and things that they built inside one of the systems. I just think like the change management around getting people to go from Harvey to Lagora or from Lagora to Harvey is really significant. And I don't think a lot of firms would do that. And And I think they're probably just happy that like their people are using AI. Specifically, the people like the innovation managers or IT managers who are in charge of this, I think are probably just like, we've got people using this.

48:08Let's not rock the boat. Like, what's the advantage of shifting? Is my rough guess, but I'm not, you may have spoken to more people about this than I have. Like, I'm not trying to go talk to firms about that, but just thinking about it, like, as I would have thought about it at Akira, like, we probably had some customers who had eBrevia or something like that, too. But I think they choose a central program and they go with that. And I think that is very much the right thing to do. Right, right.

48:41Noah Waisberg: Right. Speaking of Ilticon, when I wrote my write-up of it, I stole a line from you chatting with you at the show when you used the phrase that we have achieved peak legal tech, which I really liked. And, you know, I think that kind of— And by the way, as we kind of messaged about afterwards, I don't actually know if we have hit the peak, but it certainly is way more peak than we've ever seen before. Yeah. Right. Right. Yeah. Well, yeah. I don't think you meant to suggest we've stopped climbing. Yeah, no, no. I don't know. I'm not calling like the peak, like it's all downhill from here. Right.

49:19Because in fact, actually, like I can see there are lots of stories I could tell about how it'll continue to go up. There's ones I can tell definitely about how it won't. But I think there's lots of reason for optimism, right? Like in the vast scheme of how much law firms spend on stuff, like, and how much law firms make, like the most outrageous numbers in law firm land are not their tech spend, right? Like it's incredible how much money law firms are making. And I think it makes sense that legal tech's doing well. But yeah, Ulta was quite wild in terms of just seeing that kind of vendor spend.

50:02Noah Waisberg: I was curious, though, I mean, given this sort of state of excitement and optimism that the profession finds itself in right now or the industry finds itself in right now. Do you, as somebody who's kind of followed the world of legal AI for a long time, you feel that there are sort of important limitations to the current AI tech that the market is underestimating? Are we a little too ebullient about it or how do you see it? um yeah i think there's a bunch of different limitations number one the tech is like if you regularly use gen ai tech and i regularly use gen ai tech uh my experience has been that it makes mistakes all the time right like it just uh and and worse not only does it make mistakes it makes mistakes in ways that can be really hard to spot it'll make it'll say things in a very convincing way that might be entirely untrue.

51:03I should have a post coming out in the next week or so with some more details about a specific experience around that. So I think, number one, like the tools make mistakes. And that doesn't necessarily say that that is reduced a lot less than I would have expected over the last bunch of years. So that's issue number one. Issue number two is large law firms are generally, and that's the market I know best, is the kind of high-end big law market and those firms are generally in the business of delivering very high quality work product like 100 but it's not like actually 100 but trying to generate 100 quality work product and they're doing it on the back of these sort of fallible tools right so i think those are two issues for sure the third issue for me and like why i spent like well everybody else was like learning to play the ukulele or catching up on Netflix seasons during COVID, like I was sitting working, like I think my wife and I managed like four episodes of Succession or something like that because I was trying to like, besides just the general heavy flow of care work, I was trying to write AI for lawyers.

52:19And the reason I was writing AI for lawyers is to convince hourly billing lawyers that being more efficient was in their best interest. and actually i've had a now that i'm not really in the business of trying to sell efficiency tech to lawyers i get to have very candid conversations with them and i would say that uh the third sort of real headwind is that i still uh think a lot of lawyers are kind of unconvinced that efficiency technology helps them make more money i think right now they're sort of willing to go along with it because they think they have to do it and it's the future and maybe it's scary but what we saw with kira in 2020 and it was fine like we didn't have a terrible financial year or anything like that but when the market when the firms themselves are feeling strained they get a lot tighter on roi and we we sort of knew that that was potentially coming and you know in 2018 and 2019, especially thanks to the input of Peter Soboloff, who has been on our board since then at Insight, we were really, really, really focused on trying to show the ROI of efficiency technology.

53:32And to be honest, like, especially with five years removed, it's hard in an hourly billing firm, right? Like, what is the ROI if sort of the technology enables you to bill less hours. Like I can make cases about how it enables you to sell clients work that they wouldn't have otherwise done, that maybe it allows you to win work from other firms that aren't using technology, but that doesn't really work when everybody is kind of using the technology. You know, like it can say that it might change your realization rate, but it's a tough case. And I think it remains a tough case. And I think in a different economic environment, I would expect that my experience was that firms focus pretty hard on that.

54:17And it's a lot harder to be an efficiency tech vendor when firms are focused on that.

54:26Noah Waisberg: What about the inevitable billable hour question? I mean, it's been all of, I think, four hours since I've had this conversation with anybody because everybody's talking. It keeps coming up all the time. Yeah, I don't have to talk about it as much these days. But yeah, I think that's the basic problem, right? Like if you bill hourly, I think technology that helps you bill less hours has a mixed ROI case. Like the very best case you can make for it is it has a mixed ROI case. So then you could be like, well, the billable hour has to go away. But like one of the things that I think made us successful at Kira in our marketing was 10 years ago, 15 years ago, people were saying the billable hour has to go away like when we started kira maybe non-billable hour work was say i don't know like we'd have to check out the i guess i could put this into gpt and maybe figure it out but my guess off the top of my head is non-billable hour work in big law firms with say 18 % in 2012, 2013.

55:34And maybe it's like 25, 26 % today, right? There's a lot more capped fee work, which still like doesn't exactly count that way. But I don't think the billable hour, like it's changed a lot less over the time that I've been in legal tech than I think a lot of pundits would have forecast. and personally I like I agree it should change and I think there's a lot of uh like I'm pro change in the billable hour but I just uh my experience having been in legal tech since 2011 and Bob you longer than me has been that you know rumors of its death are greatly exaggerated yeah I've been so far 30 30 years change like I'm not I'm not pro billable hour but I just I think if you need that to change for your company to be successful, I would just be cognizant that it hasn't changed yet.

56:37It could. I think it would be great if it did. Yeah, yeah. Don't hold your breath.

56:43Noah Waisberg: So I want to ask you, so your experience in selling Kira, the pain points of going through that whole process of selling Kira helped shape your product roadmap at ZUVA. And so what happens now? Do you ever sell ZUVA or are you now here for the long haul because you just don't ever want to go through that again? Oh, no. It's like selling business is great. It's a wonder, it's like a terrible, wonderful experience, right? Like you should definitely, if it's the right time to sell your business, you should sell your business, right? Like that's just it. It's, uh, you should totally do it. It's not that bad.

57:25Like, it's not like traumatic like that. And I'd been an M &A pro before, right? Like I knew what I was getting into, our CFO knew what she was getting into. Like there were no surprises here. Like we've done this before. It's just like a very high burden. like it's a really hard professional thing to do but you should definitely do it like businesses i think uh they have a life and then they have another life after you sell like it's been really neat to watch kira post sale and just and see people from kira post sale as well some of whom stayed at latera and some of whom went off and did other things but a bunch of whom have really been thriving and i think there's just i think an individual business is kind of like a period of time and you you should try to sell it at the right point probably like i'm not uh i think there are perpetual businesses that are super cool like uh you know i'm a big fan of kushner like the thrive uh perpetual idea that there are these businesses that could go for tens or hundreds or whatever of years i don't think like a random technology company is going to be one of those and you should sell it and if you are selling you don't reach out i'll try to

58:46Noah Waisberg: help if i can but you just have to recognize yeah but like whatever it's going to be like a rough three months four months like you'll be really tired afterwards and then big deal like life goes You can get a big check. Or a small check or who knows. But you should. It's actually a great experience. It just is tough. But just because things are tough, it doesn't mean you shouldn't do them. Yeah. Our time is just about at an end here. And I always like to see if there's anything that you wanted to talk about that we haven't touched on yet. Anything else on your mind? Oh. Well, I'm sure there's a lot.

59:24Noah Waisberg: Knowing you, I'm sure there's a lot of stuff. But I think we've covered a bunch of pretty fun stuff. There's always other things. But now at this stage, I think I always enjoy talking with you. I think this time was fun. I'm glad we got to talk a little bit more on the peak legal tech thought on sort of how the market might play out. What a time, though. Like, that I think would be my main thing. Like, what a time in legal tech. What a time in legal AI. And what a time to be a lawyer, too. for me as a lawyer and as a person i hate doing stuff that just doesn't seem worthy of my time right like when literally when i was in grade school and i would get punished and had to write lines or something like that uh it wasn't like that bad like this wasn't happening all the time but i would be like i will not punch him in the playground yeah like write the lines quicker or whatever, right?

1:00:23Like, I think as someone who doesn't like doing work that I don't have to, that there's a more efficient way to do, I think this would be an incredible time to be a lawyer. I was talking to my dad's in his 80s and he was a lawyer. And he was saying, you know, if only I was a junior lawyer right now. Like, what a time to be able to just maximize your own productivity. It really feels like a pretty incredible day. And like every day there's something new, not necessarily positive either. Like I could think of several AI things that I was talking about today with a smart teammate who, which are like really negative things with AI, but I just, it's such a thought provoking, incredible time.

1:01:12Noah Waisberg: And I'm kind of glad to be working through it in an industry where I like a lot of people and getting to work on problems that I think are challenging and interesting. Yeah, yeah. If you set aside the rogue agents forming bands of - Yeah, well, I think that or the thing on Twitter today of like greater than 10 % chance that AI will kill us all. I missed that one. I'll have to go get on next. Oh, you should check that. I'll send it to you afterwards. That person actually, the person who said greater than 10%, we believe that there's a greater than 10 % chance that AI will kill us all works at Anthropic.

1:01:50I believe it is.

1:01:50Noah Waisberg: Was that the person who just left Anthropic or something? No, no, no. No, no, no. A current Anthropic employee who I believe is the head of the super alignment team at Anthropic did a tweet on top of the tweet of the guy who was leaving Anthropic saying, oh, yes, we take the threat of this very seriously. Yeah. Yeah, it's interesting that I've been in this field for 30 years and now I'm like wishing, I almost wishing I was like just starting now because I'm more excited by what's going to happen over the next 30 years. I mean, it's just crazy. It's like, so it's such an exciting time. Yeah, it's pretty incredible.

1:02:28I was in giving a talk at one of my kids' schools on AI, which is, and we were talking about just the concept that like these kids in the room were 11, I think 10, I guess, 10, 11 at the time. And you think about like their lifespans probably like assuming nothing changes in lifespan like 80 90 years right so and then i think back on just the innovations that even my grandparents saw over their 90-ish years right and it's like it's really incredible and to think what like i think the pace of change is i do think kind of accelerating and to think what what you'll get to see over the next few years, like what I'll get to see over the next year is what my kids will get to see.

1:03:18It's like really quite stunning. Anyways, I'm glad I'm kind of doing something where I get to at least be a little bit close to the coalface of using the JNI tools every day. It's quite a time. Yeah, it is quite a time.

1:03:36Noah Waisberg: Well, Noah, thanks so much for chatting with me and always good to see you and we'll have to do it again soon. I always enjoy talking to you, Bob. So anytime. Be well. Thanks a lot. Appreciate it.

1:03:53Noah Waisberg: Thanks for joining us for today's show. Hope you enjoyed it. If you'd like to share your own thoughts or comments, please go ahead and message me on LinkedIn or social media, or just email me directly, ambrosia at gmail.com. If you're a fan of Lonex, please leave us a review wherever you get your podcasts. Lonex is a production of Lonex Media. I'm your host, Bob Ambrosie. Please join us again next time for another episode of Law Next.