5 Dec 2025 · 44m · Gabe Pereyra, Sarah Guo, Elad Gil · 23 quoted facts
Gabe Pereyra argues that legal AI's next phase is not about making individual lawyers faster but about making whole teams and whole law firms more productive and profitable. Harvey sells to law firms and big in-house legal departments, and chooses to help firms become AI-first rather than building a competing law firm itself.
The central tension is that legal work is long-form and hard to verify, so it does not fit the clean right-or-wrong training signals used in coding and math. Pereyra suggests senior partners' feedback and reasoning are the real reward signal. He expects partners' roles to survive while junior work and firm structures change, and he stresses organizational design over raw model intelligence.
Early Harvey was a kind of development environment for single lawyers using GPT-4. Pereyra says the problem now is coordinating teams on client matters and making entire firms profitable, which needs orchestration, governance and security, not just smarter models.
Pereyra compares a coding environment, where agents run against unit tests, to a legal matter such as a fund formation or acquisition. Models would search document systems and data rooms and get partner feedback, mirroring how associates work for partners.
Long documents like merger agreements cannot be scored good or bad automatically. Pereyra says the real judge is a senior partner, and firms hold edit and feedback data that could train reward functions, much as outcomes years later validate good deals.
Pereyra says building a law firm and a software company are two different companies, and the Atrium experience suggested one does only one well. A firm would also be conflicted out of work, whereas supplying every firm is a far larger opportunity.
Large clients like banks and Fortune 500 companies are less standardized than law firms and often lack basic systems. Harvey's forward deployed engineers connect billing and governance systems, and their work feeds the product roadmap, a classic enterprise software approach.
Harvey builds AI software for law firms and large in-house legal teams. Pereyra says plain models hallucinate and lack a client's context, so Harvey connects them to documents and workflows, and now focuses on coordinating teams and whole firms.
Pereyra expects partners' roles to change little, since they set strategy and handle clients, much like senior engineers. He says lower-level functions will change, and firms are already discussing how to train future partners.
Code can be checked with unit tests, but a merger agreement is long text with no simple right-or-wrong score. Pereyra says a senior partner's judgment is the real test, and some outcomes only show years later.
Pereyra says running a law firm and a tech company are two very different businesses, and trying both means doing one well. He also notes a firm would be conflicted out of clients, while serving every firm is a much bigger market.
They are engineers who work inside a customer's environment to connect data and build workflows. Pereyra says big clients like banks are less standardized and ask for hands-on help, and the work also shows Harvey what to build next.
Stated by the speakers. Each quote matches the transcript, which shows what was said, not that it is true.
Pereyra said Harvey had nearly 1,000 customers and about 500 employees, and was founded just over three and a half years earlier.
“We're almost at one zero customers, 500 employees, started about just over three and a half years ago and so been kind of scaling quickly since then.”
Harvey gained access to GPT-4 when it first raised money from OpenAI.
“when we 1st raised from OpenAI, we got access to GPT four, and I think GPT three to GPT four was such a big model jump.”
Harvey announced it signed Walmart and is working with AT&T and other Fortune 500 companies and private equity firms.
“we've signed Walmart, we're working with ATT, a bunch of these Fortune 500 large private equity firms, Global 2000”
On the first day of GPT-4 access, Harvey co-founder Winston Weinberg spent 14 hours redoing his associate tasks with the model.
“Winston went in his room for 14 h and just redid a bunch of his associate tasks.”
Pereyra did reinforcement learning research while at DeepMind.
“when I was at DeepMind, a lot of the RL research I did was that.”
Gordon Moodie, a former partner at Wachtell, Lipton, Rosen & Katz, joined Harvey early and is now an advisor.
“Gordon Moody was a partner at Wachtell, which is one of the top transactional firms in the world that joined us early on and is now an advisor.”
Pereyra said the Wachtell partner worked on Michael Dell's multi-year take-private and re-listing of Dell.
“he was a part of, Uh, when Michael Dell took Dell private and then restructured it and took it public again.”
Pereyra said the Dell restructuring involved the largest debt offering of all time and a newly invented financial instrument.
“At the time, they had to do the largest debt offering of all time, they had to create, they had to invent a new financial instrument.”
Pereyra said Harvey's forward deployed engineering program resembles Sierra's agent engineering program rather than Palantir's model.
“I would say this is closer to, like, Sierra's agent engineering Um program.”
Harvey did some customization for very large customers such as PwC.
“for very large customers like PwC, we did some customization.”
Pereyra said some law firms now resell Harvey to their in-house clients and help implement workflows.
“They're starting to go and take Harvey and go to their clients and say, Hey, buy Harvey. And we'll help you build all the workflows and implement it.”
Harvey's first customer was reached through an introduction from an investor to an ex-partner, who introduced them to David Wakeling.
“Our 1st customer we actually met through. You intro'd us to an ex partner who's doing business school here, and he intro'd us to David Wakeling”
While researching, Harvey's founders spoke with 30 people from Atrium, a tech-enabled law firm.
“we actually talked to 30 people from Atrium.”
Y Combinator invested in Atrium, and its general counsel at the time later became OpenAI's general counsel, according to Pereyra.
“was the GC at Y Combinator when they did the Atrium investment.”
Pereyra estimated legal services at about $1 trillion and professional services at roughly $3 to $5 trillion.
“the scope of this, like legal, is 1 trillion, professional services is something like three to 5 trillion.”
Pereyra said Harvey grew from him and Winston Weinberg in an Airbnb to 500 people in about three and a half years.
“Winston and I in an Airbnb to 500 people in like three and a half years.”
Pereyra worked on large language models at Meta before Harvey.
“I was working on large language models at Meta.”
Elad Gil recalled that Harvey's demos worked on GPT-4 but not GPT-3.5.
“I remember, are you showing side by side, GPT three point, five versus four? And what you were doing worked on four, but not on 3.5.”
Elad Gil said Cursor launched its IDE roughly two years before the recording.
“Cursor didn't really launch its IDE until 24 months ago, something like that, probably.”
Pereyra said Harvey's initial differentiating feature was uploading a document and acting on it, with accurate citations.
“the initial feature we built that none of the products had at the time was upload a document and do something with it, right?”
Harvey hired a site lead for its New York office to scale it up.
“We just hired a site lead for New York, so starting to scale up that office.”
Pereyra said he spent the decade before Harvey trying to start a company like it.
“I actually just spent the decade before Harvey trying to start a company like Harvey.”
Pereyra said law firms are about ten times larger than before computers and the internet.
“law firms have like ten times the size compared to before computers and the internet.”
Pereyra explains the shift from an individual lawyer tool to firm-wide orchestration and profitability.
“how do you make an entire law firm working on thousands of these client matters more productive and more profitable?”
Fund formation and investment due diligence show why legal work is complex and unstructured, like a code base.
“one analogy you can think of, like understanding a code base, but the code base is all of these contracts”
Harvey's early agent-like workflows mirror how associates do work for partners, pointing toward training environments.
“you can kind of think of associates as agents, they get this task from a partner.”
Partners' roles likely persist while leverage models, training and staffing evolve practice by practice.
“how do we think of training the future generation of partners?”
Senior partners hold unpublished reasoning that models lack, and long-form legal work resists simple scoring.
“most of the value, or what you need, I think, to eventually improve these models is the decision making process.”
Harvey uses forward deployed engineers to serve large, less standardized clients and shape its roadmap.
“There is just this massive amount of work.”
The Atrium precedent and market size explain the choice to enable firms and professional services.
“I think you can only do one thing well.”
Pereyra's research background and Winston's legal insight led to starting Harvey once GPT-4 appeared.
“when we saw GPT four, I was just like, Oh, the time is now”
Pereyra argues faster individuals don't make faster organizations, so the opportunity is collaborative infrastructure.
“making someone program 20 faster doesn't make you build a product 20 faster.”
Gabe, thanks for doing this. Of course,
Yeah, thanks for coming.
Well, maybe we can just start with, like, for anyone who hasn't heard of Harvey, what is the company? can you talk about the scale and who you serve today?
at Harvey, we're building AI for law firms and large in house teams. We're almost at one zero customers, 500 employees, started about just over three and a half years ago and so been kind of scaling quickly since then. And kind of, you guys were some of our original seed investors, so yeah, good to be here.
Maybe from a most basic perspective on the product, why is it not, just, you know, copilot or Chatgpt or Claude?
Yeah, I think that's how the product started. So when we 1st raised from OpenAI, we got access to GPT four, and I think GPT three to GPT four was such a big model jump. That the intuition at the time was just give the model to lawyers and have them play with it. And I think that industry was so text heavy that you got so much value from just interacting with the models. And then I think as soon as you gave it to lawyers, you also ran into all of the sharp edges of the models.
Of they hallucinate, they're not connected to a bunch of our context. And so I would say the past, the kind of 1st two years of the company were. How do we build essentially the IDE for lawyers around these models? That connect it to all of the context you need to be productive as an individual lawyer? But I would say in the past year and going forward, the big problem we're solving is not how do you make individual lawyers more productive?
It's how do you make a team of lawyers working on a client matter more productive? And, more importantly, how do you make an entire law firm working on thousands of these client matters more productive and more profitable? And so I think when you get to that scale, a lot of the problems you're solving are not just model intelligence problems. They are these orchestration governance and kind of all of the enterprise product problems that you run into at at scale.
You've also been broadening from just law firms into enterprises, into big companies using you in concert with both their in house legal teams and external. Uh, counsel. Can you talk more about that and how that's been evolving as well?
Yeah, so we started selling to the largest law firms. And something that had started happening about a year and a half ago was these law firms started showing Harvey to their clients. And their clients both wanted to collaborate more effectively with their law firms, and they also wanted to use this directly in their in house departments. So we recently announced we've signed Walmart, we're working with ATT, a bunch of these Fortune 500 large private equity firms, Global 2000, kind of the largest consumers of legal services.
And what we're starting to build is a platform for the in house teams to do the work that they do internally. So things like contracting and this long tail of all of the legal operations you need to do that. You typically don't send out to the law firms, but also the collaborative tissue of. I'm working on a large transaction, a litigation, I need outside expertise. I wanna securely share this data with my law firm. And there's a lot of technical problems there, around security, data privacy that we wanna solve, so these law firms and their clients can collaborate effectively.
I think for, uh, you know, we have, um, a largely technical audience, but also most people don't know exactly what legal workflow looks like. Um, I think before we really started working together, I imagined, and it's just like, Well, what do you mean? I like, email my lawyer. And he thinks about it and like, reads a document and then sends something back, right? And um, there's like redlining involved somewhere, maybe there's negotiation. Can you paint a picture of just what workflow means to to you guys?
yeah, yeah. And I I think a lot of people when they think about legal, they think of consumer legal. And so I have a lease and I need to get, you know, input on that. That's completely different than what these massive law firms are doing. And so I think a really good example of what these firms are doing that. I mean, you guys will be familiar with, and I think a lot of people in the startup space will be familiar with, is what law firms do for venture capital firms or private equity firms.
And so VCs, PE firms, they do two main things you raise money and you invest it, and so the main oh, we do legal and podcasts, yeah, and podcast, which is
we do podcasts. actually important, Yeah, but there's less yeah,
legal work there. Um, and and the important things you need to do there are fund formation. So how do I structure the entity that is gonna hold all that money? And it sounds easy. But if you're a large private equity firm, you have a sovereign wealth fund that comes in and they say we need to structure it in this way because of tax implications. Then you have a pension fund that has these other requirements. And so it ends up being this incredibly complex process of how do you draft the limited partnership agreement, which can be 100 pages?
Every investor that's investing, you can have 100, and they all have side letters that modify that. And you need to understand if I modify it this way, it's gonna have these implications. And a lot of it is the project management that also goes around coordinating all of these products. If you're raising, you know, a billion dollar fund, and then once you've created that fund, there's all the investments you do out of that fund. And so, for example, when we did, you know, any of our series, you need to get a data room.
We share a bunch of data, you look at that. You need to understand the contracts we have to make sure that the revenue we say we have is actually structured the way we've claimed. And, you know, are there litigation, all these things? And so it's this massively complex process of understanding. And I I think one analogy you can think of, like understanding a code base, but the code base is all of these contracts and all this legal work.
And I think the reason legal is so difficult is the workflows aren't structured. So the same way with programming, it's really hard until these models to build tools for programmers. You basically just had an ID and then programmers did stuff in all the different languages. But you didn't have, like, Oh, here's a tool for Python, here's a tool for C. Legal's kind of the same way. And so I think a lot of why you're seeing traction in programming and legal is.
I think there's a lot of analogies of of these workflows of. They're so text heavy and the workflows. Until you had these models, you couldn't structure them in the way. I think you can now.
So one of the directions that people are going on the coding side is to build things that are being called like agentic, and it's very early on in terms of what agentic means. But basically being able to deconstruct a logic tree in terms of what are the set of actions that you need to take in a certain situation. And then having the Um AI agent go back and sort of check each one of those items, do it, go on to the next item, double check it against the prior one.
Do you do that from a legal perspective, or is that something that's a little bit more in the future relative to where code is today?
yeah, we're, we're starting to do this now. And actually, like when I was at DeepMind, a lot of the RL research I did was that. And so when we 1st got access to GPT four, we had the very strong intuition of okay. You're gonna be able to, you know, string a bunch of these model calls, or eventually do things like reasoning models where the full agent is differentiable. And even the 1st day we got access to GPT four.
Winston went in his room for 14 h and just redid a bunch of his associate tasks. And when I looked at the work he was doing, it was essentially like this hacky agentic, where he said, Okay, I would need to go. Look up this case law, summarize it, take that summary, use it to draft. And so seeing him do that gave us the intuition very early on of that's the direction this is going. And you can kind of think of associates as agents, they get this task from a partner.
That's, hey, I have this high level case strategy. I wanna see if I can find a bunch of case law that supports it. Can you go research that, look it up, cite it, write me a memo? And so a lot of the systems we're starting to build look a lot like that. And I I think one interesting direction that the coding labs, the research labs are going, is building these RL environments where you deploy these agents.
And they can interact with a code base and see if they can pass unit tests, and in legal, that RL environment is a client matter. So you have all of the context of a fund formation, an acquisition, a litigation, and the models are starting to learn. Let me go in the document management system and see if I can find this, go in the data room or do case law research, get feedback from the partner. And so I think that research direction is is super, super interesting.
It's it's really interesting. You make the associate analogy, cause I remember, um, when I led your guys's Series B, which I think was maybe two years ago now, it's a while ago. I called a lot of your big customers and talked to the head of the law firm or talked to the head of, you know, um, some of these institutions. And one of the things that I thought was really striking is number one. They were adopting legal software, which before was really hard to sell into them, and because of what you were doing, being so striking and important, they were adopting you really fast.
The 2nd is, um, they weren't, uh, threatened by it, and I w I thought, oh, they'd be threatened. Because it may augment or eventually replace certain aspects of law, et cetera, or, or, you know, help sort of change that dramatically. And one of the insights they kept bringing up that I thought was really interesting is, they said, as we think ahead, as this sort of AI tooling and agentic workflows spread through Harvey and companies like you, um, how do you think about the future of a law firm?
Because instead of hiring 100 associates, of which you assume ten will be partners, eventually, maybe you only need 50, maybe you only need 20. And so are you even hiring enough people to know who'd be a great partner? Because you're, you're gonna shrink the set of people that are needed to do certain tasks over time, right? Right now, that isn't true, it's augmentation, it's expanding business, but that could happen in the long run. How do you think about the future of law, or what law firms will look like, or, you know, the evolution of all that?
yeah, this, this is a great question. I think it's changed a lot in the past couple years. I think something we are starting to talk with law firms a lot is, how do we think of training the future generation of partners? Where, to your point, these law firms have these leverage ratios, where you have a lot of associates, but much less partners. And there is value to that because not everyone is gonna become a partner. And part of going through that process is how you find, oh, this is the person that I would trust to do this very complex acquisition because they've gone through that experience.
And so I think the part I'm optimistic about is if I think about over ten years ago when I learned to program, it was super painful, right? Like you had to go on Stack Overflow. It was hard to learn multiple languages because you're like, Okay, I'm just gonna like, learn Python, I'm gonna learn TensorFlow or something. It was just like hard to even learn that. It was hard to ask questions. When I was at Google, you don't wanna ask a bunch of questions, cause people will be like, Oh, you don't know that.
And They get I think,
stuck all the time.
yeah, exactly. And now with the models, it's like programming is so fun to learn. Because you can just be like, here's how to write this in Python, translate it. Why is it written this way? And you can learn this so much more quickly. And we see lawyers doing that with Harvey, where they'll say, generate this merger agreement. Why did we structure it that way? And so we're already starting to see some of that. But I think the really big opportunity for law firms is how do they take all of the internal partner feedback and data that they've created and use that to start training?
I think that's one big piece. I think another conversation we're having is, to your point, how do you just generally start restructuring firms? I think this is one where we have some intuitions, but a lot of it is going to depend on the firm, the region, the size, their specialty, the types of clients they serve. I think one of the things that's very challenging with law firms is they are really a collection of all these practice areas.
And so the firms that specialize in litigation look different than the firms that specialize in large transactions versus like mid sized transactions. And usually the big firms do a collection of these. And so a lot of what we're spending time is practice area by practice area. Can we go and sit with? Here's the fund formation group and their private equity clients and start thinking about what that would look like in terms of the workflows, the staffing, the pricing?
Uh, and I think it is a really interesting problem, where a lot of the value in the product and the platform is not just the product itself. But how do we help enable these firms to transform? And so when you think about it from that perspective, like, our goal is, how do we make these law firms more profitable? And it's not just a product problem, it's thinking about their holistic business and where do we fit in?
In kind of that bigger picture?
hm, hm. Yeah, it's really interesting, cause when you look at the set of functions that a partner fills, and I'm, I'm thinking in particular of consulting firms, unless about law firms, simply because I'm a little bit more, uh, familiar with consultancies, uh. Some of it's the pattern recognition, the high level thinking, the strategy, and then part of it is like the sales and really being able to make that client connection. And and so it's interesting to, to your point, think about more broadly, how can AI augment all parts of their business versus just the the legal workflows?
yeah, and I, I, to your point, I don't think that part changes. Where it's like, when we think of the like, we're now larger consumers of legal services, and when we think of the best partners we've worked with. I don't think the models are doing what they do anytime soon. And I think what's interesting is, I think the role of law firm partners actually doesn't change that much. In the same way, I don't think the role of very senior engineers changes with this.
Because you're largely delegating work and what you're getting paid to do is here's the high level strategy, here's the right abstractions. Go write the code or write, do the legal research to help me do it, and I will interface with the client. And so I think that my guess is that doesn't change too much, but some of the like lower level functions do change because of this technology.
One of the things that you said that I thought was interesting, like in another conversation, um, that we were having was that. There's an analogy that you can make between, like a a great senior partner, or like a Gordon, Moody type, as like a distinguished engineer working on systems at Google, right? Um, I think for a more technical audience, or just general business audience that doesn't really know what Gordon Moody does. They might assume what a lodsiter is like, isn't like 50 of that, like his network, right, or his reputation.
Um, but you know, what you were pointing out is like, there's expertise in the ability to predict, like a sequence of arguments. That is going to, like, get you to the answer you want or manage risk. Um, how does that? How does that translate to an RL environment or a task for you?
yeah, this is a good question. So for maybe for background context for the audience. Gordon Moody was a partner at Wachtell, which is one of the top transactional firms in the world that joined us early on and is now an advisor. And kind of the analogy I was giving is, why is a senior, distinguished, uh, distributed systems engineer at Google so valuable? And a lot of it is the experience they have architecting these systems. That none of this is public, this won't, won't go into the models for a long time.
And so if you're building search at Google, these people can just point out, hey, if you build this system this way at this scale, it's gonna collapse for some reason. That is super, not intuitive. And one of the examples that Gordon talked about early on was he was a part of, Uh, when Michael Dell took Dell private and then restructured it and took it public again. And this was like a multi year, super complex financial and legal restructuring of an incredibly large business.
And what he when you talk with him is incredibly good at. It's the same feeling as when you talk with a very senior engineer where he can. Just. He has the whole picture of this legal entity in his head. At the time, they had to do the largest debt offering of all time, they had to create, they had to invent a new financial instrument. And so it's just understanding. If I need to raise this much money to do this part of the transaction, this is how I would structure it.
And so a lot of the value he brings is not just the relationship, but it's just that technical understanding of how you architect these things, the same way that how you architect like very large software projects. And I think when that translates to an RL environment, part of what is missing from the public models is the process of looking at one of these entities and figuring out, given all of the context of, I wanna do this merger, this is the right way to structure it, just that process.
And a That's a reasoning lot.
trace, right? Like, for an expert, just like it would be in code, yeah. And if right? Yeah,
you looked at that data set for one of those transactions, it would be. Client comes to Gordon, says, I wanna do this large merger or acquisition. And then there would be meetings and emails talking about, Okay, this is the background of the two companies. This is roughly how we would structure them. These are all the things we need to look into. And a lot of the data would be Gordon giving these tasks to associates to say, Okay, look into these risk factors of similar transactions we've done.
They would do research and say, Okay, maybe we could structure it this way. And then he would point out this really subtle thing that, hey, actually, in this case, if you structured it this way, this thing's gonna happen. But none of that shows up, like all you get from these public mergers is like an SEC filing. And so you do see the final result. But most of the value, or what you need, I think, to eventually improve these models is the decision making process.
The same way you need these reasoning traces to to train these models to do kind of any of these reasoning tasks.
One of the, as you mentioned, like, the labs are all very focused on, um, uh, RL scaling in, um, like coding and math domains. Um, I think those is like, highly verifiable, right? Not perfectly so, but like, how, how do you think about the appropriateness of like law for RL, given it's not as easily verifiable?
yeah, this, this is one of the biggest problems. And I remember we had conversations early on when we were trying to figure out what is the right evaluation structure. So I think the hardest thing about legal is most of these tasks are very long form text generation. And so there are definitely subsets of legal work that are super verifiable. Of. Go in this data room and just find all the change of control provisions that you can kind of build these traditional data sets, but for something like generate this merger agreement, it's really hard to just give some binary like this is good or this is bad.
And I think this has been like a big research problem, like with all the labs we work with. And also internally, there is just this open question of how do you build that reward function? And if you think of what that reward function is at the law firms, it's the partners, right? Like, at the end of the day, there is no way to verify this. Besides the senior partner who's done a bunch of these said, Yeah, this looks pretty good.
And so internally, these law firms have a bunch of data of. Here's all the edits that went into this and the feedback. And so we are starting to think about, how do you use that to train these reward functions? But I would say that is one of the really big problems. But I think one of the interesting things is, I think you actually have the same problem in programming, where I think in the short term programming is verifiable, where you can look at unit tests, but once you get into real software engineering, like the unit, there is no unit test.
It's like, I deployed So, like system design this, yeah, it's like I at scale, yeah, deployed this and a million users used it for six months and it didn't crash. And so you get, and and it's like, mergers are the same, where it's like you can make sure the filing is correct. But at the end of the day, it's like, three years later, the companies are still merged. And they didn't take on litigation they didn't expect, or something like that.
And it's like, that is eventually the really valuable human experience, right? That's what you pay really good software engineers, or really good lawyers, where they have that decade long track record of they can build these systems and they haven't fallen apart. And a lot of these stuff, the same way, you can't unit test, it's like hard to verify. So it is, I think this really interesting open research problem.
One thing that you guys are doing on sort of the UM other end, from pushing the bounds of, uh, what Harvey products and, um, the models can do is like, just get them deployed. And, uh, you recently started this for deployed engineering force. Um, this is confusing to to me because I'm like, well, you're not necessarily like an application building company, which is how people have traditionally thought of. FTE like, why are you doing this?
I would say the model that we're operating under is not a full palantir. Let's go into the code base and kind of build custom software. Uh, I, I would say this is closer to, like, Sierra's agent engineering Um program. But what we're starting to run into a lot is, I think early on, we did a really good job of building a horizontal platform. We did not that much customization for customers in the sense of building specific things for specific customers.
I think the thing that was nice about legal was we could build things like workflow builder and things into the product that would let customers customize the product. And then for very large customers like PwC, we did some customization. But now we're getting to the point where when we're starting to talk with law firms about, hey, we wanna take a bunch of this data and help you build a model or build agents. There is some amount of we need to go in this environment, into your environment and figure out how to connect all the data.
We're starting to connect to a lot of their like business systems, so their billing systems, governance systems. And then, especially when we start working with the Walmarts, the very large banks, the Fortune 500, they're much less standardized than these law firms. And so there is just this massive amount of work. Where we go to a large bank and they say we don't have any document management system for our legal department, can you just build us one?
And there is a massive amount of demand of. We just want smart technical people to sit here and help us think about our business and our operations and how we should start mapping that into Gen AI systems. And for us, it's a really good way to figure out the roadmap. Where, for example, Glueal is like one of the fastest growing private equity firms that, uh, we recently started working with. And they we meet with them all the time.
And they're just like, there's all these things that we feel like we could map into Gen AI. We don't quite know what it's gonna look like, but let's just sit together and figure it out. And so I would say, that's a lot of the genesis of the of the program. Of. How do we just get more people that can work with all these customers and start kind of paving the way of some of these new roadmaps in different verticals?
yeah. I think what you're describing, too, is a very standard enterprise playbook. And I think in Silicon Valley, people almost forgot because of the SaaS era. That if you're Oracle, if you're Dell, if you're IBM, if you're any of these larger organizations, this is how you sell software, right? You have a platform, you have a bunch of customization around it. People have bespoke data sets. They may not always have the ability internally or enough people to implement certain connectors or systems.
And this is like the standard way to do it. And then, as you do it over and over again, you start repeatedly turning that into part of the platform. And so
right. And and I I think a lot of these started with doing something that resembled FD. And then you get big enough that you get this like implementation ecosystem. There's all these 3rd parties that will come in and implement
they'll be like the certified like vendor, whatever, yeah,
And and I think the interesting thing we're actually starting to see is law firms are starting to do this for their in house clients. So they're starting to go and take Harvey and go to their clients and say, Hey, buy Harvey. And we'll help you build all the workflows and implement it. Because we have the scale and the expertise to build this where typically these in house teams,
that's really cool.
the smaller ones don't have, like the budget or the in house to build this. And so, yeah, I think there is kind of And a lot
that seems like that could be a good revenue driver for the law firms that you work with in terms of a new line of business that they can offer. yeah, and and Yeah, it's really
some of them are starting to think about it that way.
I was really struck by. Uh, it wasn't. I don't know if it was day zero. You can correct me, but it was within the 1st year, where the very 1st version of Harvey was really an individual lawyer productivity tool, right? Uh, I'm an associate at a more senior personal law firm. I wanna get a piece of work done. Can you just make it less painful? Uh? But the transition quickly to like, how do we transform the business, make the business more profitable, like organize teams being the ecosystem, I think happened, happened pretty quickly.
And like anything that is a business transformation, just requires, like, you know, a lot of engagement. Um, and so I, I, I w. Given how much you guys have invested in customer success and how that's like driven adoption. I feel like a big piece of it is just how quickly AI has happened, right? I? I would not necessarily have predicted that all the customers you're working with would be like, yes, in year one and two of this company selling, we're adopting.
Um, but, uh, you know, part of it is you guys are helping them, right?
yeah, no, and I I think this was like, when I look back, it was still surprising how quickly some of these law firms adopted this. Like our 1st customer we actually met through. You intro'd us to an ex partner who's doing business school here, and he intro'd us to David Wakeling at a. And O. That was in our 1st year, and they went from a small pilot to firm wide and investing in this. And I think, I, I mean, I think you're seeing this in a couple verticals with, like, cursor, open evidence, where this technology is so transformative for these industries that just are so text heavy and knowledge based.
They just haven't had tools like this. That. I think early on we did find these customers that were like, Oh, this is worth really betting on. But yeah, I think the pace has still been pretty surprising.
I asked the internet through X, Uh, what questions we should ask you? Um, and a popular one was like, Why aren't you guys building a law firm? are you gonna build a law firm and compete with all your customers?
Yeah, no, we, we get this question. And I mean, I think when we 1st started Harvey and we were doing research, we actually talked to 30 people from Atrium. And I think also interestingly, Sam and, Uh, Jason, who was the GC of OpenAI at the time and was the GC at Y Combinator when they did the Atrium investment. What struck us was the people who worked there said it was a really good idea and they were super excited about the prospects.
And then there were some challenges around the legal and the execution. But when we dug more into it, the big challenge that they ran into was. You're essentially just building two different companies, right? You're building a law firm and you're building a tech company, and it's already really hard to, like, build product, engineering, do AI, scale sales. And I think the big issue you run into, if you try to do both of these is, I think you can only do one thing well.
And doing a law firm well is very different than building a software company well. I think that's one point. The bigger point is for us, it feels like the best outcome is if we can figure out, how do we make every law firm? how do we help every law firm become an AI first law firm? not how do we build one ourselves? And I think the real problem we're trying to solve is, can we make every law firm more profitable?
And a part of that is how they work with their clients. And can you make their clients get better, faster, cheaper legal services? And I think solving that equation at scale is a much bigger opportunity than if you build a single law firm. Because you get conflicted out, you can't scale this. And so I think this is probably like, this is something we don't do, but we've gotten this question. But yeah, I I think it's kind of not the focus for the company.
Analogous to other, uh, markets in software, um, law feels like an area where I've been very surprised personally about how how large the scope of the problem is. Eventually, if you're really ambitious about what you can do. Um, I didn't realize, like you, you were telling me, you know, if you do a really large MA, let's say of like two global companies, Microsoft Activision or something. You're like, there's 100 outside counsel firms here, you know why?
Because in New Zealand, where both companies have customers, you have like a tax implication. And the dude who understands that lives in New Zealand, right?
oh, it's CR. Yeah, it's crazy.
And and so I I think, like, you know, like other markets like the SMB version of this looks really different from the like, high end enterprise version of this. Um, and so I, I do think it's like, hard, it just seems hard to imagine, like coalescing all of that expertise in a law firm and a software company at the same time. Versus, I'm like, well, Harvey now has like, what 40 customers in New Zealand,
exact, yeah. or is it? Yeah, And and I mean, if you think about those transactions, it's also not just law firms, so there's investment banks and you maybe have PWC or a tax advisor. And there could be an HR consultancy that helps you think about how you're merging headcount. And so for us, the bigger opportunity seems, how do we build the platform that lets professional service providers and their clients collaborate? And I think a lot of the problems you need to solve there are like, the biggest is the secure collaboration across many of these entities, the secure data sharing.
How do you build and deploy AI systems across these very complex projects? And I think, to your point, the scope of this, like legal, is 1 trillion, professional services is something like three to 5 trillion. There's just this massive amount of room to grow. And we think our expertise is going to be in building the product, the technical systems, the AI systems that enable that. And we want to give that infrastructure to all of the different law firms rather than compete with them.
Because I just don't think you can.
I think one thing that's really striking, um, about this sort of wave or era of AI, is that there's deeply technical people building giant companies in really different industries. And you come from a research background, you worked at one of the major labs. In terms of foundation models and other areas, RL environments, reinforcement learning. What has been your biggest surprise in terms of transitioning into being a founder and running a company and building something from the ground up like that?
they may be, maybe not surprise, but biggest, like mental model shift is, I think, the ten years before Harvey. I was doing a mix of mainly AI research and then trying to start companies, but always largely as an IC. And I think the shift from this started working and scaling, just how much I had to change my mental model of the type of company we're building, how you do this at scale, how you operate. I think that was the biggest surprise, or like thing that I've had to had to change.
But it's been a crazy experience, kind of going from, you know, Winston and I in an Airbnb to 500 people in like three and a half years. And then I think also how you build these products at scale and kind of the complexity of this industry like. That has been like a really hard but interesting experience.
It's been amazing to see what you all have accomplished, you know, it's. It's such a short period of time. And
I, I was thinking back to, like, when we pitched to both of you guys, like three and a half years ago, and we were like, Hey, AI, plot for legal, and you guys were like, sounds good. But we like, I mean, we had some of these ideas, but I think they've really like,
I think a really important aspect of that, too, is, um, you all started this company, um, before GPT four came out and before a lot of the shifts in the models happened. And so I remember, are you showing side by side, GPT three point, five versus four? And what you were doing worked on four, but not on 3.5. And you were part of that very early wave that had conviction. This was so important, it was a trend.
Was that because of your experience in the labs? Was it something else, like, what drove you? Because not many people were actually starting AI companies when you all got started, and it was kind of just, to your point, AI plus legal, like nobody was doing that.
yeah, yeah. It's something where now everyone's like, Oh, this is such an obvious idea. But It's a text at the time, yeah, in this way. Yeah, every, yeah, now, yeah, text in, text out. But at the time, yeah, no one was thinking about this. I think it was a combination of a couple things. A lot of the best people at the time I had worked with had gone to OpenAI. And so I was working on large language models at Meta.
You saw GPT one, GPT two, GPT three. If you were working in AI for the past ten years, kind of one of the big problems were, how do you pull all this together? Cause you built systems where, okay, this is really good at vision, this is really good at like specific things. But no one really had the like general solution. And you saw, saw, you saw things like Lambda. And I think with that trend, like what I'd seen is anytime you kind of make that initial okay, this is how you do it, you can usually just scale and this stuff keeps working.
And so with three, three, five, we were like, this is getting really interesting, but it's not quite there. And so the bet was okay. OpenAI may be one of the people to crack it. Like, I know a lot of the people there, uh, that was part, and then I think the other big part was just Winston, uh, was a lawyer. And I think we had become super close, never thought we'd start a company together. But just the way I heard him talk about the legal industry, like he, even though he was a 1st year associate, just had this intuition of not just the work he was doing, but the structure of the firm.
Like, I would hear him talk about the firm and be like, here's what all the different partners are doing, here's like, why our firm's strategy is this way he had. He was in the process of convincing some partners to leave to start a law firm with him, which is insane for like. And so it was just like, Okay, this will be really fun. He'd showed me a bunch of his legal tech. It was like this seems like the perfect application.
And then when we saw GPT four, I was just like, Oh, the time is now like this is the perfect like application.
I think it is like, really noteworthy. Uh, where I was actually, like, you know, even six months into, um, working with you guys, being like, Well, are capabilities really gonna advance that quickly? And both you and Winston were like, absolutely right. Like we should, we should have the ambition to, um, take on the full complexity of, like any type of legal work that's possible. Because the models will keep getting better. And like that seems like an like, a like, a super obvious mainstream point of view today.
But in, I don't know, the middle of 2022, I think it was like, a, um, it was a strong, unique intuition to have.
yeah, yeah, I th. I think that was something we did really well, where we just had this belief. Where I think the same that you see with the programming products, where if you had built something where it's like, all this does is like, check that your Python code doesn't have bugs, which you could have done better with three five. Like you wouldn't have built something like Cursor. And the intuition was just these models can help you do any programming task in any programming language.
And I think we felt that same way in legal. And I had a bit of intuition. I did like a bit of investment banking, private equity, and it was the same workflows where you could just do any of them with these models. And so I think keeping the product open ended enough that it gave us the room that now we can build into all these things. And and like professional, other professional services, I think that was super important.
It's a it's a really interesting analogy because for code, it took an extra two years, I think, for the main coding companies to really emerge as the ones that are likely to win, right? Um, and so you folks started, I think, um, three and a half years ago, and you had a product almost immediately and you were up and running really fast. And then I think, um, Cursor didn't really launch its IDE until 24 months ago, something like that, probably.
yeah, that sounds, yeah, And then cognition was slightly in that era, and then obviously, cloud code.
S six months later, yeah,
And so everything kind of came kind of in a time delayed way for code. Even though GitHub Copilot was one of the 1st products and everybody knew that that was really important. And I always think that's really interesting because there were so many coding companies that got started under the premise. But somehow it's these ones that started a little bit later that really were the ones who took off. And so I always wonder, why is that?
You know, what caused that shift?
well, I, uh, my guess is my part of my intuition here was just you guys were like, let's say like, very, um, uh. AGI is less trendy as a word now, but very capability pilled from the beginning, right? So both you, Winston and you, as a shred of an investment banker turned AI researcher, were both like, it's gonna be able to do so much
Yeah, I think a lot of
of this.
the coding people thought that too. yeah, they were, yeah, I I think people super.
were, I think people were a little less ambitious, like three years ago, at Yeah, there are, least some set.
there are some people who are gonna go and build AI models, and, you know, I actually feel like, uh, people are very ambitious. I just think that maybe you folks immediately focused on product and that was part of the difference.
I think it was finding the right form factor, and I think in legal it was maybe a bit more obvious where the initial form factor was. Essentially like, like the the initial feature we built that none of the products had at the time was upload a document and do something with it, right? And that is a lot of legal task. And it was that, and then do really accurate citations. And when we showed people that, they were like, Oh, this is crazy, cause that's so much of my job.
I think with coding, the initial models were also not quite as good. That you needed maybe a bit more capabilities of the base models, and then you needed. I think, figuring out the right way to, like, integrate this into the into
into the IDE. the ID. But Yeah,
I mean, I, I remember where it's like the 1st version of the product that I built. It was mainly like, I used GPT four because, like most of my background, was like distributed systems and AI research. And I was te i, I still don't know. React. It's just like raw JavaScript and kind of like putting this together. But I'd be like, Hey, GPT four, like, help me make this. And like, you could kind of already see it at the time with programming, and that was part of what gave me the like intuition that I could analogize to, like, what Winston was doing.
hm, hm. What, uh, you mentioned that you folks have gone, uh, from basically the two founders to 500 people over the last three and a half years or so. You're obviously growing really quickly. The business is working, you know, you have tons of customer demand. Uh, what are you hiring for? what are you looking for in terms of the next set of employees, or what types of roles are you hiring for right now?
Yeah, on the on the technical side. So we mentioned FDEs, I think looking in general, like across roles of just strong engineers. And then I would say, maybe specific call outs. We just hired a site lead for New York, so starting to scale up that office. Uh, more folks on kind of front end and scaling product in general, and then more AI folks as well. So but yeah, anyone strong engineer like, please apply.
okay. Last question for you.
How many pull ups can you do? Just kidding,
We did find out, Yeah, well, I. I don't know if that was the max, but these guys can both do 15 pull ups. Guys with a wink and we a smile. can do more. could okay, okay, guys, we get it. Um, um,
You mean, in one set?
in one set, in one set of
okay, we gotta do the three, yes, 24 h challenge.
Yeah, oh, what's that? Just how many how many can
can you do?
you do in 24?
No way.
it's a lot.
yeah, you can upload it to your TikTok
Oh, that's cool.
from from the No Priors TikTok. Does Elon have TikTok? No, I Okay, I just don't know, wanted to start well, then I don't have that changes a TikTok. Yeah,
Do you know what TikTok is very good for? Arawan videos? oh, super good. yeah, Yeah, that's really they're good, good.
it's really good.
Yeah, there's some really funny ones, Arawan videos yeah, like,
like the LA. yeah, like Spoofer smoothie thing.
Arawan, or like people like, like, there's one where it's, um, the Miami girl visits Arawan Oh my god, and she's like, Arawa, and they're like, Are you a fun? So it's very good. I highly recommend not have that I know I in writing. we now used to do a couple. have, Yeah, no, that
thanks a lot. that, and Twitter, I feel like, is where my has the time goes,
I guess. TikTok, Arawan videos,
Okay, you can pick, we'll pick one of these two. I asked some other people involved in the company s What question should I ask? Gabe? oh boy, We covered some of them, but one of them was like, why do you still sleep on an air mattress?
okay. So I don't sleep on an air mattress, I have a good mattress, I don't have. A bed frame is where that's coming from. And so what happened is when we moved from LA. to SF. My bed frame broke. And then the 1st year and a half of the startup, things were so crazy. I was like, this is what a startup founder should do. And at some point I was like, I need to get a bed frame.
And I ordered one and it came. And I got a call from the apartment and they were like, Hey, you didn't sign out, you didn't fill out the insurance, like the rent, the mover's insurance. So we can't let them bring this up. And I was like, Okay, I called them. I was like, Hey, do you guys have renter's insurance? They're like, we're UPS, we don't, they're just like, we don't do that. And then I was just like, I don't have time to deal with it and I haven't dealt And you just with it.
So
gave him like a delivery
I just decided we have other date. Okay, problems to solve. So yeah,
So physically can't do anything except the company
just only.
right now. Yeah, and pull ups,
e exactly.
yeah. Other question was, Um, you know, there's a bunch of foresight in starting Harvey when you guys did. When you look forward, do you have a prediction that you think others don't necessarily agree with you right now? That is not mainstream.
So I. One comment I'll definitely make on the foresight is, I think a lot of like. We've gotten comments of like, oh, overnight success and oh, you saw this coming. And I would say I actually just spent the decade before Harvey trying to start a company like Harvey. So I think it was just I was super early. And then eventually it was like, Oh, now's the right time, and then you were kind of in the right position.
I W My guess is, I think people now are catching up to how capability pilled, as you called it, like Winston and I were. I think people in Silicon Valley, I think people have a good sense of where these models are going. But I think generally people don't appreciate how much better they're gonna continue getting. It's hard to internalize, It's really weird. Yeah, it's really weird. And and like I think
build things. And I'm like, Oh my god, code gen works, it just really works now.
it's crazy, yeah. And to me, it's like, I think the interesting thing will be the transition from like. These models are really smart individually. But if you think about, like a lot of what we've done in the past 20 years with SaaS, it's how do we use software to make these massive organizations? And I think that will be the continued trend, where a lot of what we're starting to think about is like, law firms have like ten times the size compared to before computers and the internet.
And I think that's gonna happen again, but in like, maybe a different way than the past 20 years. But I think that to me, like a lot of people still talk about co pilots and individual productivity, and I think a lot of the the things we're starting to think about is like organizational productivity. And like, how do you build these systems at scale, where both for our internal engineering team? Like, I think a really interesting question for the for the cursors, the codexes is like, making someone program 20 faster doesn't make you build a product 20 faster.
And so starting to think about, like, what is the broader infrastructure you need so these companies can develop software and product faster? And then kind of same analogy to legal, I think that's kind of one of the things we're thinking about that I maybe don't hear people talk about as much.
kind of collaborative AI, in some sense, it's. It's sort of like the Figma transition of your, uh, individual contributor designer versus working collaboratively with a design team. And what you're talking about is doing that for law, doing that for code, doing that for different verticals and having AI as a layer on top of that. So it's super interesting.
yeah, and I I think to that point, it's like, how do? How are humans and AIs going to work super effectively? Cause even at these large companies, you have huge teams of different specialized people that have different functions. And I think when I hear a lot of people talk about these models, they kind of talk about it as like, oh, AI will just get smarter and do all of this. And I don't think that's the way this evolves.
The same way. It's not just like, hire 100 people and now you've built Walmart. It's like, so much of it is like Million how you, um, three, yeah, 3 million, actually, yeah, how you organize all of these. And I think that will be like one of the really interesting problems yeah, super for these.
interesting. Yeah, I'm. I'm seeing that a lot in the context of both AI driven roll ups as well as this company, Bringo that help get up and running. Where a lot of the AI implementation issues are around people, management, workflow optimization, it's it's much less about can you build the AI, and much more how do you actually change the organization to be able to adopt it properly? So
yeah, no. And we're starting to work with a lot of private equity firms that I think it's interesting, like starting to see how they're thinking about that, cause I think that will be like, really interesting space.
awesome. Thanks, Gabe.
thanks for coming.
Thanks so much for having Yes, I am me.
Mr. Jim plubs, but.
Find us on Twitter at No Priors Pod, subscribe to our YouTube channel. If you wanna see our faces, follow the show on Apple Podcasts, Spotify or wherever you listen. That way, you get a new episode every week. And sign up for emails or find transcripts for every episode at no dash priors dot com.