Eve raises to bring AI agents to law firms, targeting the $2T legal market

Well, thank you. All guys, thank you so much for coming on breaking it down. I'm sure we'll have you back on again very soon. We'll talk to you soon. Congrats. Our next guest is in the re room waiting room already. Uh legal ai fundraising announcement. Let's bring in Jay from Eve. Jay, how you doing? What's happening?

I'm doing great. Thanks for having me on, guys. Busy week for you. Give us the news. Give us an introduction. Give us an explanation of the company. Give us an update on the news. Absolutely. And I can't I can't wait for what's coming next. Oh yeah, you know, you know, get that mallet ready. But we are Eve.

We're AI for plaintiff law and we just raised $103 million at a billion. Boom. There we go. Um I fantastic. You know, we we've talked to various players in in AI legal.

uh we don't you know follow the the space nearly as close as as probably you know many of the investors do and you do but uh I first time hearing about you guys was yesterday. Give us the backstory on the company because you're not obviously just coming out of nowhere. I'm sure you guys have been cooking for a while.

No, absolutely. So we we took an unconventional approach which is you know we basically took time to understand what plaintiff and defense law means and how they're different. uh which maybe a lot of people a lot of people don't know.

Uh but plaintiff lawyers, you know, these are the ones that you would typically find fighting on on behalf of individuals, right? So they typically don't charge billable hours. They make money when they win for you. Think labor and employment issues, personal injury issues, you know, family law, things like that.

Um on the defense side, you get the nice benefit of retaining a company and kind of cashing it in every year, right? and then you have to typically they charge by billable hours. So on the two sides there's very very different business problems and opportunities.

Um on the plaintiff side they care about how do I represent more customers with the best of my ability to win as much as possible for them right either changing conditions or uh money right in civil litigation.

And on the defense side, it's like how do I do more for the client, you know, for the same amount of time or maybe more time and delivering value.

So on the plaintiff side, what we found is there's actually a really good merger between AI and the efficiencies you get from AI and their business model because you're able to go deeper and actually benefit more clients. Is that because you're charging on a value basis, right?

you're taking a fee based on the value that you capture for the client or the or whatever like you know if there sorry the plaintiff if there's a settlement whereas on the defense side you're charging billable hours and you're tech you know there's some you can do a calculation as a client and say like how much value am I getting do I feel like this is a good deal but it's less of a because it's based on billable hours it's less of a it's less like value based pricing yeah you you nailed it right so we we typically charge based on how many clients they represent.

So there's a very clear alignment of incentives between what we provide for our customers and what they provide for their clients. So we're kind of in plaintiff law they're you're incentivized to use more and more and more and more AI.

So you can have less time to serve each individual client and generate more revenue which is yeah again maybe different than on the on the defense side where it's like oh well I I'm like AI because it makes my job easier but I don't want to be perfectly efficient unless you're just have overwhelming demand.

You nailed it and this is the really cool part about right now is our customers are using this to level the playing field and flip it in a way that's never been happening before. Right. So, if you think about the average insurance company, they employ so many lawyers.

And when you have a personal injury attorney showing up, it's oftentimes one person in court versus like 10 or 20. They just have endless resource to run you down, right? And I mean, you know, typically if insurance paid out, like the industry wouldn't exist at all.

Um, so, you know, there's that's what they're there to equalize and fight for on your behalf. Um, and what people are doing with Eve is actually using it to now act as though they have a million-dollar attorney backing every single step of the way from intake to resolve resolution of anyone. Right?

So now you can actually do all these complicated legal workflows specific to plaintiff law and we go deeper into these workflows than really any other tool can and deliver that to the high level of bar that you would want it to have as a client, right?

You don't want as a client to worry about how many cases are they working on? do they have time to spend on me? Uh, and you know, the law firm has the same problem. How far away are we from just AI on AI just violence in in every case?

It feels it feels like Sorry, I I'll let you answer, but I I've heard a story like 10 years ago about uh a law firm that set up a automated system just using not AI, just normal software to run through the line items that they were billing their client on the corporate side, uh and detect what the client would say, hey, we're not paying for that.

We're not paying for your two-hour lunch. Like, so don't write that in. And so anyone who was writing their billable hours would know that it got flagged. they'll take it out. And then the and then the then the other party got software as well to detect what was going to be, you know, maybe debatable in the billing.

And so they did have this like software on software fight in legal specifically, but just at the billing level. But now we're we're going one layer up. But I'd love to hear your answer on how it all plays out.

You know, this is actually really interesting question and one we're seeing play out live is again the value alignment incentives. So on the plaintiff side, people are using it and they're coming up with tactics because they know the other side has to bill hours. Yeah. Right.

And the other side doesn't want to really reduce the hours quite yet. So what's happening is, you know, one of the common ways defense slows you down is they kind of start discovery requests and responses.

These sometimes take up 48 hours for plaintiff law firms to go fill out and our customers using Eve are able to come back and return back right away. Right. And then the other side is like what w you know what is this going on?

Uh and they want to spend the time to uh to respond and that's actually raising the settlement value of firms. Mhm. Because you're basically eating into what the lawyers would be charging their clients right on the other side. Interesting. What are what are some of the downstream impacts of what you're building?

you know, assume that you assume that you can get to, you know, 50%, you know, or some meaning double-digit percent of of of plaintiff law firms using Eve or comparable software if there is a comparison. Uh what what are you know, is that going to impact the insurance industry?

Like are they going to have to you know, if if if every uh plaintiff has this super lawyer, you know, what what what are kind of downstream effects? Great question. There's a few different downstream effects I foresee happening and we're seeing a little bit of this already.

Um, so for example, let's take labor and employment, right? So think you get fired, harassed, etc. You're getting representation in in this industry. Typically a lawyer want to see at least $5,000 in fee before even representing you because again they're doing it for free until they win.

So they're baking that into how they think about it. And what we're seeing with EVE is actually a lot more proono cases or them going down the value chain to take on cases they otherwise would never have. Right. And now more lawsuits. We're getting more lawsuits. Yeah. So are lawsuits going to go exponential?

Are we going to see exponential? I mean this would be very American if if we used AI to just 10x the amount of lawsuits annually. But it feels like Great great question. What's interesting and maybe most people don't know this is very few cases actually get to a lawsuit perspective. Y right.

So in personal injury about 70% or so of cases actually get resolved before a like at the demand letter stage. Exactly. They would get resolved demand letter stage or some form of mediation before might even just be like hey yeah like you didn't pay this invoice.

You know we we but we both know that this could turn into a lawsuit very quickly with AI. So we're just going to settle quicker and we're going to get to like the median outcome where both people are happy. That's exactly because once a lawsuit starts, you know, $700 an hour accounting hours. It's arbitration, right?

Arbitration essentially is like what's the maybe the long-term result of a lot of like these AI battling it out. What's uh give us some numbers. How I mean you raised it a billion so things uh are seemingly pretty good, but what what kind of traction have you been seeing so far?

How many times have you been one of the one of the interesting bits is uh usage as well and we've been fortunate I mean one of the really crazy things about this industry is I was just in a conference probably three weeks ago and I asked this question every year last year I was there I asked how many people have used AI every day or you know in the last week 10 people out of 200 raised their hand this time around personal injury attorneys 90% of them raise their hands wow right so the amount amount of awareness and traction it's seeing is crazy.

But what they've always been scared about is like how do I use it? And this is kind of the problem we're solving, right? Because if you're just using chat GPT or Gemini or whatever, you're spending a lot of time fixing things at the end and that's not good enough for lawyers. They lose trust very quickly.

So we had to go Yeah. If you have an associate that just constantly makes every single time they make, you know, one or two mistakes, those associates don't, you know, eventually they get turned out, they get let go, right? So you want to hold the AI to a higher bar. Yeah. Exactly.

And then like when we hit that, you know, extra nine of accuracy, which I'm sure you two are very familiar with. Oh yeah. Our usage and adoption rate kind of blew up, right? So we went from I think around 13 customers or so start of last year to now more than 450. I think we're approaching 500 now. That's good.

Um and and uh you know dollar signs are increasing very rapidly as well which is helps us invest even more behind R&D which I'm super excited about. The cognition team just rewrote all of Devon on the back of uh Claude 4. 5.

Are you in a similar world where a where a new model can cause like a fundamental shift in how you think about building the product? Are you sort of model agnostic? uh how closely are you tracking what the foundation labs are doing?

This is one of the key decisions we made early on in the architecture is we worked very closely actually funnily enough we were one of the first productionized apps I think on cloud 2 back in the day uh and they actually expanded context window just for us. Wow. So you have multiple modality of increases right.

So model capability gets better. Y context window gets better. You know, agent functionality gets better. All of these have a pretty drastic implication on how deep into a workflow can we actually go. Yep. And what gets enabled.

So somewhere somewhere out there there's a there's a defense attorney who isn't online, not using AI, just doing everything the oldfashioned way, and he's probably just getting his world rocked like month over month. Like, how did you respond to that in an hour?

I was expecting I was expecting that to be I was expecting that to be three weeks. That's great. Uh, walk me through this this thesis. Tell me if I'm right or wrong. Um, Enthropic's been incredibly focused on the flywheel of code, coding agents.

They're hoovering up tons of data reinforcement learning on top of the the code, developing a really dominant code model. Uh, OpenAI has been less focused on API, but chatbt is dominant. I've talked to I bet if you pulled those lawyers, a lot of them are using chatbt.

If they're not on the enterprise plan, Chat GPT and OpenAI are going to be able to train on that data. And so, you might have a data flywheel that actually makes OpenAI's APIs better for legal work. Is that a reasonable thesis? Well, I mean, according to the agreements we signed with them, they're not supposed to. Okay.

So, we do Well, well, that's because you're on the enterprise plan. What I'm talking about is like if I a person don't have an agreement with chatbt and I get uh, you know, a contract for an advertiser or something and I just uploaded that to ChatGBT, they're going to train on that.

And and I'm not saying that's your your company. I'm saying me as an individual, I'm just putting extra data in chatbt. They're getting more data. It makes the model better and then you benefit on the API side even though they're not training on your data. Yeah.

I think one of the interesting bits for us is we benefit from the general training open all these providers do. Yeah. Mostly because actually law is weirdly touches everything. Yeah. You know when you get into a lawsuit sometimes it's medical, sometimes it's technology, right?

Sometimes it's like quantum computing or something and these lawyers have to somehow be up to speed on these things, right? And the general training actually really helps. So that's one part of data flywheel.

There's another part that I don't think people are thinking enough about which is the one we benefit from in how we build Eve which is all these intermediate tasks right are actually not captured by anyone.

An example of this is like you actually can't even go Google how do I write a really good you know demand letter for a moving vehicle accident for again state farm or something you know it just doesn't exist and what they're doing internally is they're coming up with multiple artifacts to produce that final output right so everyone has a workflow produce this document this document this document and pull information from these three to make this final document and they've all come up with their own workflows and one of the things that we've been I been fortunate to do in Eve is we actually allow them to kind of teach Eve to do their workflows kind of like how they teach a new associate or a new hire and then after that point it just follows it you know diligently without fail uh at a high level of bar right this doesn't miss things like a person would because it finds everything instantly um so that's one of the ways in which it's changing and maybe I'll be a little spicy like what I kind of predict will happen here is I actually think these law firms will will basically transform into software companies over time, right?

So, think about early Amazon or something. Everyone started, you know, being able to offload their uh hard disk and all these things into APIs.

And what is happening with Eve is now instead of relying purely on human labor to scale, you can start offloading these very crucial tasks that were previously not APIable into EVE, right? And you can view yourself as now, hey, I'm not just servicing one client one-on-one anymore.

I'm figuring out a way to service everyone that I'm an expert in, right? Right? So, if I'm an expert in dealing with traumatic brain injuries, I'm going to take on every single traumatic brain injury case and do the best I can for them and be a rock star at it.

And this is like the bene main benefit that software companies benefit from. They've consolidated their R&D efforts to make a really good product and sell it at, you know, incrementally lower additional cost for a net new customer. I think it's super exciting. I think the best the best lawyers are uh your friends.

They're expensive friends, right? But they they help you, you know, they're just they're they're there to think through problems. And I think that's just going to continue to be I think that's going to continue to be valuable for a long time, especially for the market that's paying for legal services today.

But they're going to be able to be a one person. You're going to have a one person hund00 million revenue, you know, plaintiff law firm, I imagine, right?

because you just have this mass massive efficiency and one person who's just great at um yeah uh the what what's uh the doctor's you know bedside manners basically like guiding somebody through a you know potentially