How AI Is Reshaping a Senior-Lawyer Law Firm (Full Transcript)

[00:00:00] Speaker 1: In the last two or three years, we've done no marketing, but yet we were able to grow 40 to 50 percent a year just with the efficiencies that we were gaining with AI. We don't have any younger lawyers. This point doesn't fit our model. With AI, we have the equivalent of an army of associates. It's constantly available 24-7.

[00:00:28] Speaker 2: Hi, I'm Stephanie.

[00:00:30] Speaker 3: And I'm Zach, and welcome to The Lawyer's Podcast, one of the many ways that we help lawyers build healthier firms, better businesses, and more sustainable lives. Today, I talk with Brian Belt about the shape of AI in a, honestly, a really AI-forward law firm that is functioning, I mean, on the bleeding edge of how you can use AI in your firm.

[00:00:54] Speaker 2: We've been working with Brian for a while. He's doing some pretty cool stuff, so I'm excited to hear this one.

[00:00:59] Speaker 3: Yeah, he really is. Well, Stephanie, we just got back from ClioCon, and right before ClioCon, they have the American Legal Technology Awards, and we've got some news on that.

[00:01:15] Speaker 2: I know. We're so excited. Look, guys. We're winners.

[00:01:19] Speaker 3: Yeah.

[00:01:20] Speaker 2: It's like the invisible award, because it's glass. It's very pretty, but it does not photograph well.

[00:01:27] Speaker 3: It does not photograph well. Yeah. So, we found out a little bit ago that we were nominated for the journalism category, Stephanie and I. And at the awards, we were, you know, we actually got the award for this year's American Legal Technology Journalism category, which, I mean, it's, you and I were talking, it's a surprise. Like, it's really cool. It's really cool. And honestly, I didn't realize how cool it was going to be. I know.

[00:01:56] Speaker 2: I think we were both kind of surprised. We're like, wait. This is way cooler than we realized. And maybe it was a little bit of like, I don't want to get my hopes up and then not win this thing. I don't want to get my hopes up and then not win this thing. So, you know, I kind of told everyone, oh, it's no big deal. And, you know, we sit here and talk to each other and we think there are people out there listening to us over airways right now. And we don't always really know. We don't know. Do you hate us? Are you throwing things at your radio dial right now?

[00:02:37] Speaker 3: Yeah. Are you rage listening to us? Which, I mean, okay.

[00:02:41] Speaker 2: Yeah.

[00:02:41] Speaker 3: It's fine.

[00:02:42] Speaker 2: It's a thing. You know? Zach, why'd you say that again? Ah, Stephanie, stop. Stop with that. I mean, what do you, what do you people think of us? But at least there's a few people in this world that think we must be doing something okay because they gave us an award.

[00:02:57] Speaker 3: Yeah. Yeah.

[00:02:58] Speaker 2: Well, and we're, you know, like we said, we're, you know, we're, you know, we're doing this for a reason. Yeah.

[00:05:21] Speaker 3: All right. Well, now, without further ado, here is my conversation with Brian Belt.

[00:05:26] Speaker 1: Hi, I'm Brian Belt. I'm a founding partner of a very small transactional law firm in Miami, Florida. It is composed entirely of senior lawyers, all with ex-large law firm experience. We don't have any younger lawyers at this point. It doesn't fit our model. And we're leveraging our AI and our legal tech to take the place of the younger lawyers, and we hope a little more efficiently.

[00:05:55] Speaker 3: Brian, thanks for being with me. I really appreciate that actual, that intro right there. This is why a lot of times we let the guests introduce themselves, because that was a fantastic intro into what we're kind of talking about in this episode. You and I have kind of talked a lot at various places, various conferences and things like that about how you're using AI in your firm. And I thought you'd be a great guest to describe kind of the shape of what it looks like in a firm that is, what I would actually say that you guys are AI enabled, as opposed to just using AI. Talk to me a little bit about how have you incorporated AI? Well, I guess first, how big is your firm? What are we talking about?

[00:06:45] Speaker 1: We're currently looking for a fourth and fifth lawyer. We had a person that we're replacing. We've got a number of people talking to us. We're looking for a bilingual corporate lawyer, and we're looking for another commercial real estate lawyer. I think we've spent a lot of time over the last three years growing just internally and we were able to do that because of the AI. We for the last two or three years, we've done no marketing. My wife was sick, getting some cancer treatments for a period of time. So we didn't even go out and have lunch to people with people, but yet we were able to grow 40 to 50% a year just with the efficiencies that we were gaining with the AI. We've now reached the point where we feel like our platform is built out the way we want it to, although you're never, with the way AI is changing every day, you never really reach that point. You're in a process of continuing improvement. So I think by April of next year, we'll probably be six or seven lawyers that will just be working with the existing contacts, the existing clients, and referrals. We're mostly word of mouth. So we built this out essentially, I think it's important to have an internal structure and internal incentives that are consistent with the use of AI. And by being senior lawyer only, if you don't use AI in our firm, you don't succeed. And we have the right incentives. I think we're moving more and more towards a system that is closely aligned with the client's interests, which I think is a point of a lot of discussion these days. And there's numerous facets to that. So a day in our life, our tech stack basically consists of Harvey, which is becoming more and more by the day, the central place where we work. It includes iManage. We converted from iManage, or from NetDocuments to iManage approximately a year ago, and it works well for us and it's very compatible with Harvey. We're also using DeepJudge, which is a indexed search system that again, integrates with Harvey and iManage very well. So what we've spent our time doing, and this I think leads into how we're using it right now. In the last year, we spent time building a larger database. I think we had a pretty good one based on the collective number of years that we've all worked for large law firms. We've spent our time improving access to that database and cleaning the database. There's no lawyers that I think enter, for example, enter the information on profiles, document profiles, that's really necessary to create good metadata. iManage now has an enriched feature that we use that helps us with that, and DeepJudge compensates for a lack of less than perfect data hygiene, if you will, and Harvey helps. So we've really spent our time doing that. And then we've also spent our time mapping processes, taking four or five or six similar deals and seeing how we do those. What are the common processes? What are the common friction points? How can we do things better, faster, cheaper? Some of that has to do with our internal law firm economics. So now we've done all that, and we'll continue to do that, and we'll probably do that until the day we stop, but now we're at a point where we're really seeing a lot of results from it. And we start off, obviously, with client intake, and that is enhanced from the use of AI, gathering the data that we need to make client data smooth. We have developed a supplement to our CRM system where every day we look at, we have the Harvey agents look at new contacts, new documents, and draw as much contact information out of that to supplement our system, including using it to help us categorize contacts. And we divided our contacts and our CRM system into segments. There's about 49 primary segments and sub-segments. For example, we may have a segment for commercial brokers, and then we may have sub-segments for office brokers, industrial brokers, et cetera. So that's our starting point. One of the key innovations of Harvey is there's a brilliant integration with iManage, and that integration has continued to be improved upon and strengthened every month. Essentially now, iManage data, for a particular matter, syncs with our vaults in Harvey. So we're always up to date with the information in our system. And so we build the vault.

[00:12:46] Speaker 3: Let me take us a couple of steps back, because I think we've gotten into explaining some of the nitty-gritty to people before we've really talked about kind of the structure and the shape, broadly, of how you're using and how you're thinking about AI in the firm. One of the things that you said early on is that your firm is taking AI and, I guess, first thing, growing, size-wise, people-wise. You're adding people. I think one of the interesting ways that you're adding people, though, is that you are looking at this as senior attorneys. You're adding senior attorneys. Why is it that you're able to do that instead of bringing in, I guess, younger attorneys and saying, hey, use our AI to make yourself better? Why that structure instead of trying to kind of put the AI on top of maybe a newer attorney?

[00:13:47] Speaker 1: People talk about, oh, the technology is expensive. It's only expensive if you pour it on top of an existing model that isn't really geared towards that. We think the best use of AI is by senior lawyers. Senior lawyers know the questions to ask. They know when they're not getting back what they want to get back. There was a recent Google study out, or DeepMind study out, that talked about when senior lawyers are given AI, and then it's taken away from them, they actually still produce better results than before they had the AI. Why? Why do you think that is? I think there's some kind of foundational elements where AI feeds curiosity. It changes the way you think. It's a thought partner sitting over your shoulder that's available 100% of the time. On the other hand, giving a young attorney AI without structure and without guidance is like giving a 10-year-old a chainsaw to play with. It doesn't produce good results, and it doesn't enhance their learning. I think there's ways to enhance that learning. The other part of it is the weakness of young lawyers is they're people. People have certain weaknesses. When you're in a large law firm environment, as I was for more than three decades, there's this feedback loop that repeats itself between a partner or a senior person and a younger lawyer. You'll get a document from the younger lawyer. It won't be quite right. You'll give them feedback. You'll tell them to go back and do it again. If you're lucky, you only repeat that process twice. Immediately, you've created, assuming you could produce an equal product doing that, an equal quality product doing that, you've created a delay. That delay, assuming your associate is always available to work on your matter, that delay can be two days or four days or a week. With AI, we have the equivalent of an army of associates that's constantly available 24-7. The availability of AI versus the availability of a young lawyer, I think, is important.

[00:16:19] Speaker 3: Let's talk about how to structure that. In this system of creating the law firm, you're trying to supplant or take over a little bit of the job of maybe the younger attorneys, the associates, things like that. It sounds like, though, that this comes from structured information, like structured data. I think that's what I gleaned from the first part of this conversation when you were talking about how using Harvey and Deep Judge and iManage to put metadata on everything, to organize everything. What's the importance of having that structure inside of your firm, or SOPs or ways of operating inside your firm?

[00:17:15] Speaker 1: That's the starting point for our structure. Good data and easy access to it obviously enhances the ability of Harvey to get us where we want to get. Beyond that, Harvey's done a lot, especially over the last four months, to become an orchestration platform. Inside that platform, you can build processes. Your share spaces, for example, we're building out templated share spaces for every kind of matter that we work on. Within that share space will be the workflows and the agents and even standardized tasks, which Harvey now allows us to do, that will be useful to our attorneys for that particular type of work. They don't have to go hunting around. They don't have to look and guess what's the best agent. I think a corollary of that is if you're going to be effective using the AI, you should take some time to underpin it with agents and workflows that you've built. That's where the process mapping comes in, because with the process mapping, you can say, here's what I'm actually doing on this deal. I can say, I want an agent to do this, this, and this, and here's opportunities for automation. Here's in terms of document automation, which now I think has reached a revolutionary state in the last month, here's opportunities where we can build better documents and react quicker. It's not just typing in a prompt. You're building an AI infrastructure.

[00:19:11] Speaker 3: I'm trying to tie this back into the concept of having really enabling senior attorneys as opposed to trying to make junior attorneys smarter. What I mean by that is a lot of times I think people think of using these AI products as a way to have more access to information, to be smarter than they are, to be more effective, but what you're doing is actually coming at it from a different perspective and saying, how do I use AI to enable my very intelligent, very experienced attorney to do the things that they do already more quickly and more effectively?

[00:20:00] Speaker 1: I think you've hit on a key point here, and what we're really trying to do is apply senior judgment, build a system which can effectively apply senior judgment to using the right agents, querying the AI appropriately, building AI out in a way that is appropriate, not repeating processes. One of the things that we've found is, and I was surprised by this, is as a result, if you use AI with senior lawyers and you enhance speed of your output, the speed itself results in better quality, and there's a direct relationship between speed and quality.

[00:20:56] Speaker 3: The speed that they're able to go at helps to increase the quality of the output that they ultimately have.

[00:21:05] Speaker 1: Enormously. Fascinating.

[00:21:09] Speaker 3: Okay.

[00:21:10] Speaker 1: When you're not consumed, lawyers, you know this, lawyers are always facing a time gun to their head, whether it's responding to other parties, whether it's responding to your clients, or whether you're facing budgetary constraints that don't allow you to spend an infinite amount of time on a particular transaction. It's like a two-minute drill in football. You want to start on the hosing team's 20-yard line. You don't want to start on your own goal line. So when we start, we know we're always facing time constraints directly or indirectly, but we're starting farther down the field, and what that allows us to do is to ask questions that we couldn't ask before, because A, we didn't have time to ask the questions, we didn't have time to think, or getting the answers was going to take too long. It also allows you to do things that, and you don't really think about this until you see the benefits, but it allows us to do work that we couldn't possibly do before because there isn't time to do it.

[00:22:23] Speaker 3: I love that, because I think that's the thing that we forget about a lot of times, is like, yes, we can do things faster, but we only think about how much that reduces the amount of hours that it takes for us to do it, and therefore we extrapolate out into, well, that reduces the amount that I can bill. But what you're saying is that I can do this significantly better because I have the time and the capacity to do the extra things that I would have thought, eh, there's not really the ROI on that, or it's not in my client's incentive to do that. So talk to me about how does that play into the relationship with the client and our billing relationship there?

[00:23:05] Speaker 1: I think you have to look at some intangibles to get the answer to that. One of the tangible things, though, is we're landing work when we're on the other side of big firms that we normally wouldn't get. I can tell you just one deal that we landed because of the AI, paid our whole tech staff for the year three times over. We got that work because it was a complicated transaction, several hundred documents, five AMLA 50 firms involved, 80 people on the email list, and the firms all, you know, I think there's a, we'll get to the firm adoption versus attorney adoption issue, but we were able to manage that document process with just one lawyer very smoothly and very quickly, and the client saw that we can do it. I think part of the client relationship issue here is explaining to them what you're doing, how much time you've invested in this, and being very transparent about it. If you want to value bill or raise your billing rate, client's got to understand why.

[00:24:18] Speaker 3: Yeah, and that's kind of the opposite way that people's, not the opposite way that people are going, but the opposite way that I think people's knee-jerk reaction goes, because they go, instead of, can I tell the attorney how much better I'm doing for them by using AI, a lot of times I think people are saying, do I really have to tell the, not the attorney, but the client, do I really have to tell the client that I'm doing this faster because of AI, because I don't want them to know that, you know, this is quote unquote easier, and what you're saying is, we're doing it better, and I want to shout it to them that we're doing it better. I want them to know.

[00:24:54] Speaker 1: The other way we, an intangible form of measuring how we're doing with this is we're seeing a much greater volume of referrals from large firms that we've worked across the table from, and on a lot of our deals, you know, we took our big firm practice and we somehow, I had these delusions of grandeur, and that I was going to be able to somehow do it, and the AI and the process has actually, it's been a therapist, it's satisfied those delusions of grandeur, and it's allowed us to do the things that we wanted to do. But when we're sitting across the table, probably 60 or 70% of the time from large law firms, and then after we're done, we're getting work from them, that's a way of measuring, I think, the quality and the speed and the way that we're doing work. Another way that we measure that, in addition to our clients are paying faster. We have two clients right now with invoices over 30 days. I would say our average aging is somewhere between five and 10 days. We've had clients pay a monthly bill of six figures, they were European-based, and we send out the bill at eight in the morning, which is, what, two o'clock their time, and before the end of our working day, we've got payment. It's certainly a measure of client's satisfaction, and we've noticed that our personal relationships with clients have improved. We're closer to these people, partly because AI is great, and these platforms are great at allowing you to scale up almost infinitely and instantly, but they're also great at allowing you to scale down, to handle efficiently the small matters for important clients.

[00:27:02] Speaker 3: That small matter becomes worth it to you and to them.

[00:27:07] Speaker 1: Yes. If we're doing, we don't go around looking for residential closings, for example, but we represent a lot of ultra-high net worth individuals, and when they say, I've got a $700,000 condominium purchase I want you to handle for my son, hey, no problem. It's being able to answer their small daily questions that they might have been too afraid to call us about before, and there's a totally different tone to a relationship like that than one where they call you because they've got a major transaction, and regretfully, they've got to use a lawyer.

[00:27:49] Speaker 3: Yeah, I think that's a good way of saying it. Regretfully, they've got to use a lawyer. Yes, they trust you. They would rather go to you than anybody else, but that does shift it to, okay, well, I know this person is doing this as quickly as it can be done and doing it extremely well, and I want to go to them. That is a different scenario, yeah.

[00:28:11] Speaker 1: I think there's also what's really come about in the last 45 days, and we're seeing this with Harvey. There's a number of good platforms out there that are a bit like communism. They sound good, but practically speaking, it's very difficult to use them. For example, task management systems, transaction management systems, CRMs, they all suffer from common pain points, and that is getting the information in quickly and efficiently into the system. Now, and I think Libora is probably doing the same thing. Now, what it's allowing, what AI and Harvey is allowing us to do is to embed the AI in those platforms like DealClosure, like Avoca, like Moople, and allow you to automate the ingestion of that information and to make suggestions as to how you keep those systems up to date, how you exploit them, and triggering tasks for you. So suddenly, these other, I wouldn't call them platforms, but these other applications become much more useful, and you don't need a junior lawyer or a paralegal or legal assistant to help you ingest that information. To manage all that.

[00:29:42] Speaker 3: Yeah.

[00:29:43] Speaker 1: And yeah. And so the term headless now is becoming a term of the day, and I don't think we'll ever go completely headless, but we're going to go headless up until a point it asks us, do you want to do this? Or here's our suggested output. We'll always be able to have the last say, but that's another reason that you don't, that senior lawyers work better for us. They can pull the trigger on that increased volume of information that you're, or the output that you're receiving as a result of the AI, they can pull the trigger more quickly, and they can make better decisions as to how they want to modify it, reject it, or accept it.

[00:30:29] Speaker 3: Okay. This feels like a good stopping point, even though I know that we can go for a long time on describing a lot of these things. What would you tell somebody that is looking at this and saying, okay, Brian, you've been doing this a while. It sounds like you have this very integrated, you've been working on this. What would you tell them should be their first step if they're overwhelmed with all this?

[00:30:58] Speaker 1: I think the first step to me is to buy a turnkey platform. This is just my opinion, and a lot of people will vehemently disagree. You've got things like Claude that are very good, but Claude comes with words like build, construct, connect, and security. Yes. I think it starts with a platform that's intuitive. There's a narrative out there that these turnkey platforms are expensive. They're not. They're actually, I think, dirt cheap. I think even we spend, we've got a very, I think, robust platform. We're spending about 5% of our total gross on tech. It's never been cheaper to start a law firm. So, I would say, don't be afraid. Don't listen to the narrative that it hallucinates. Hallucinations are like someone driving four times over the legal limit of alcohol, a car with no brakes, and habitually running red lights. And then you say, oh my God, I got an accident. And then I would say, go to some of the trade shows. You might want to start off with the ABA tech show. But some of these more robust tech platforms are not at the ABA tech show. They're at Legal Week. They're at Iltipon. And I would encourage people to build close relationships with their vendors. They will actually listen to the small firms. Harvey talks to us all the time. But the race decision is, you've got to go out and watch some YouTube videos on tech that you're considering. Read through their support site. And by the time you get them on the phone, you understand the platform. And then, you know, change management isn't what change management used to be. I don't think that term should even be used anymore. That was a useful term when you were, you know, enhancing Outlook a little bit or adopting a new billing system. This is transformational. It's lifestyle. It's a completely different way of doing things. It's almost like you have a different job. So you have to approach it that way. And if you want to sell it to your people, do some Netflix trailers and show them, here's what you can do. One minute. It's hard to capture a lawyer's attention. But Netflix captures people's attention with one minute trailers. You can do the same thing. You have to market this and have lawyers begging you, when are you going to roll this out? Uh, and the last comment on the change management would be take some, take an extra month and build out some agents, build out some workflows. It's really easy. If you can type or talk, you can build a workflow or an agent. It's that easy.

[00:33:59] Speaker 3: I love that as a, as a quote right there. If you can type or talk, you can build a workflow or an agent. You're absolutely right. Yeah.

[00:34:05] Speaker 1: And, um, you know, start with, I would start with the turnkey platform and then build into it. You'll, you will see what you can do. I think you need a quality document management system, whether that's net documents or I managed, that's the, you know, where all your information is gathered and it's, and, and they built those out so that they're accessible to Harvey or accessible to Laura and they're accessible to other apps. And I think the overwhelming quality that you need is you need a sense that you don't have to do things the same way you've always done them. You don't have to do the grunt work. What you have, what you have to do is ask the question. I wonder if, or could I do this? And the answer will almost always now be yes.

[00:34:59] Speaker 3: I love that. And I love that as a quote to, to end on Brian, Brian, thank you so much for, for your, uh, your thoughts and your, your ideas in this. I really appreciate you sharing how you're operating, how your firm is operating, how you, how you think about this, um, going forward in a, you know, AI enabled or enhanced law firms.

[00:35:20] Speaker 1: Well, well, thank you, Zach, for the opportunity. And I love sharing about this because, um, AI makes the practice of law joyful and, and that's a different thing for an old guy like me. Um, so thank you, Zach.

[00:35:37] Speaker 3: Love it. Love it. Thanks, Brian. We'll talk to you next time.