Maximum Lawyer

Watch the YouTube version of this episode HERE


Michael McCready has spent years building McCready Law across seven states, and now he runs the firm from Puerto Rico. In this episode, he joins Tyson to unpack one of his most interesting systems: a digital twin trained on decades of his podcasts, presentations, SOPs, employee materials, deposition transcripts, opening statements, and closing arguments. His team can ask, “What would Michael do?” and get an answer rooted in his management style, firm culture, values, and legal thinking.


Michael also breaks down how his firm uses internal LLMs and AI agents across intake, case opening, 30-day, 90-day, and 180-day reviews, medical records, demands, litigation, and more, all with human oversight and audit trails. He and Tyson discuss why law firms are becoming technology companies, how owners become bottlenecks, what “delegate, don’t abdicate” looks like in practice, and how a firm can start using AI without building an advanced tech stack on day one. They also talk about how Michael built a 180-person, 24-lawyer firm that can operate across seven states while he lives in Puerto Rico.


What You'll Learn
  • How Michael trained a digital twin using podcasts, presentations, SOPs, employee materials, and legal transcripts.
  • Why data permissions, guardrails, human oversight, and audit trails matter when AI touches law firm work.
  • How internal LLMs can turn SOPs and employee handbook material into direct, usable answers for a team.
  • How AI agents can review intake decisions and case files throughout the lifecycle of a matter.
  • Why law firm owners often become the pinch point, and what “delegate, don’t abdicate” looks like in practice.
  • How Michael built a seven-state law firm that he can continue leading while living in Puerto Rico.

Highlights 
  • 00:01  Meet Michael McCready and his AI digital twin
  • 02:39  How the “What Would Michael Do?” digital twin started
  • 07:37  Internal LLMs, data permissions, guardrails, and human oversight
  • 16:24  Why law firms are becoming technology companies
  • 21:13  Training AI on voice, legal thinking, and trial work
  • 24:26  Getting the owner out of the way without abdicating responsibility
  • 29:54  AI agents across the case lifecycle and practical places to start
  • 44:26  How Michael runs a seven-state firm from Puerto Rico

Connect with Michael
Website
LinkedIn


🎟️ Get your MaxLawCon tickets: maxlawcon.com
🔍 Vet your vendors: beccaslist.co


Maximum Lawyer helps law firm owners build businesses, not jobs.


Resources:

Creators and Guests

Host
Tyson Mutrux
Tyson is the founder of Mutrux Firm Injury Lawyers and the co-founder of Maximum Lawyer.

What is Maximum Lawyer?

Maximum Lawyer is the podcast for law firm owners who want to scale with intention and build a business that works for their life.

Hosted by Tyson Mutrux, each weekly episode features candid conversations with law firm owners, business experts, and industry leaders sharing real strategies and lessons learned in the trenches.

If you're ready to grow your firm with less stress and more support, this is your next must listen. Subscribe today.

Tyson Mutrux (00:01)
Today's guest runs a multi-state law firm, but he lives in Puerto Rico. And over the last year, he's built a digital twin. His team can literally ask AI, what would Michael do, and get answers based on his thinking, his decisions, and the way he's run his firm over the last 30 years. We're also going to talk about what it actually looks like to run a successful law firm while living somewhere completely different, how Michael made that transition, and whether he thinks this is where law firm ownership is headed over the next decade. Michael, welcome to the show.
Michael P. McCready (00:34)
Hey, good to see you again, Tyson. How are you
Tyson Mutrux (00:36)
Good to see you. I give people a little bit of context because I know you fairly well. Sometimes I forget that, right? I know the guest, and the listener doesn't necessarily. So give people a little bit just the high points about you and your practice.
Michael P. McCready (00:54)
Yeah, sure. I'm Michael McCready. I have a law firm that started in Chicago, McCready Law. We are now in seven states. We're in Illinois, Indiana, Missouri, Wisconsin, Iowa, Michigan, and Minnesota. Single event personal injury practice. I started the firm 28 years ago. I haven't handled cases in about five years, and that's really been what's been responsible for the growth of my firm. I know that you've talked endlessly about this, about you know, on up with other with other guests about you know working on your business and not in your business. And about two years ago, I relocated to Puerto Rico and I run my law firms from here. So that's a high-level overview of where I've been for the last quarter century.
Tyson Mutrux (01:46)
Yeah, I first met you. It was at John Fisher's first mastermind. And I don't I think it might have been twenty twelve, twenty eleven, something like that. I can't remember. Yeah, long time ago. So
Michael P. McCready (01:54)
It's about twelve, thirteen years now, so do the math, probably around then.
Tyson Mutrux (01:59)
It's just been really cool because I feel like you've always been ahead of the game when it comes to technology. You were I can't remember the software you used to use. I don't know if you still use it, but yeah, you
Michael P. McCready (02:08)
Yeah, we still do. Smart Advocate. It's been it's been responsible for our growth.
Tyson Mutrux (02:13)
Yeah, you've been a massive advocate of Smart Advocate. Yeah. So it's and I will tell you, even though I've never used Smart Advocate, I actually w I really like the people over there. They've always been really, really s respectful and I've always I've always heard really good things. I've actually not ever heard a complaint about Smart Advocate. This is not a promo for Smart Advocate, but
Michael P. McCready (02:30)
Yeah, no, no, no. And we won't name names of other case
Tyson Mutrux (02:33)
Sure.
Michael P. McCready (02:34)
Management programs where they you can't speak as highly of them.
Tyson Mutrux (02:38)
That's exactly right. Yes. But let's so I want to t I want to get intoday's topic because it's interesting because your team can literally ask what would Michael do? Because like you've made a digital clone of yourself, which I think is just so fascinating. Like whose idea was this? I want to know that. Whose idea was this?
Michael P. McCready (02:56)
Well it was actually mine. Let me backtrack a little bit about technology. So one of our firm values is technology. So we've always embraced it. And when AI started coming on the market, I mean we've been using it for almost three years now. So what a lot of people think of AI today is what we've been doing for a couple years, right? It's the chat back and forth, and whether whichever LLM, large language. Model you use, it's very similar. You do a prompt and an answer comes out and you know you can have a continued conversation. So the f so that's where it all started and I think most people are largely familiar with that these days. You know where AI is now and where my firm is you know we're implementing AI agents. That's a conversation for another day. But just to give some people a framework. So, with any LLM, it depends on what you train it with, okay, what you feed it with. And so the big ones, you know, ChatGPT and Claude Anthropic, Grok, Gemini, a couple others, they have been ingesting, you know. All the information possible to put into their to their LLM, their large language model, to give it context to be able to answer questions. So once you understand that, then the next leap isn't it's not much of a leap is to create your own database, your you know program your own LLM and feed it with that information which is used to you know to give responses. And that was what the impetus of all this was.
Tyson Mutrux (04:42)
What was the problem you were trying to solve inside your firm that made you say, Hey, we need to build this?
Michael P. McCready (04:48)
I wouldn't even say it was a problem, it was a possibility. Right? And so with AI, so much of it is you're limited only by your creativity. So if you're still experimenting with you know with using AI and chat and so forth and so on, you know, the first skill that you really need to learn is how to frame your query. So you know, back in law school, we learned the Boolean search, right? This within four words of that, and if you were Good at the Boolean search, you could get cases. Now I sucked at the Boolean search, right? So, you know, I was never very proficient at Westlaw or Lexis. And that's why I loved when Google came along and it's largely natural language. So you type in a question and boom, you know, Google gives you a Google doesn't give you an answer. Okay, Google gives you a URL or a website that its algorithm believes is responsive to your question. Now Google's made a lot of money by having an algorithm that usually gives you the website for your answer. But contrast that. That is not the answer. The difference with AI is it's not linear and it doesn't give you a URL. It processes that kind of information and gives you a synthesized answer. So Each of those, the Boolean search, the natural language, and now AI, both require or all require different skills in crafting your query in order to use the tool most effectively. So once you master the query, you're limited only by your imagination. So that's where this idea came from. Is that you know, what could we feed into an LLM that would allow us to query it and give it a framework or a context to provide answers.
Tyson Mutrux (06:50)
So we built a an AI native case management system. We call it Foxy. And our AI that's inside of it we call Roxy. So we can ask Roxy any question that we want. And it's to the point where people even call Roxy her, she. Like it's funny how like whenever they're referencing her, they I say it still, but they say her. It's just funny how it's almost like it's a human. I'm curious, have you named your twin? Is it have a name?
Michael P. McCready (07:15)
No, it's just you know there's it's WWJD, what would Jesus do? So it's W D. So what would Michael McCready do? So that's kind of what we you know colloquially refer to it. But no, it's you know it's just another resource where I think that you know and We built this out, but it's the technology is becoming more and more readily available that you know I think that the future of most law firms is law firms will have an internal LLM. Law firms will have multiple LLMs. When you start talking about feeding information into an LLM, one important thing you need to take into consideration is access to the data. So when I first found out about ChatGPT, you know, I was going hog wild and I said to myself, you know, let's just put everything in my law firm for 25 years into this LLM. What a great idea. No, not a good idea, because there's things in there that some people in the firm shouldn't have access to. When I if I had gone Down that road, it literally you could have asked the LLM, what does Michael McCready think of my work ethic? I don't want somebody to do that, to have access to that. So you know, data you know. Fencing off data and permissions, you know that becomes much more important. And then guardrails. So you know LLMs you know, have human have to have human oversight and human training to allow them to do things and not do things. So you know, you know, ChatGPT and Claude, you know, there are certainly questions which you can't ask it, right? You know, I haven't tried it, but you know, we can all imagine what those are. Those parameters are programmed by humans, those are human. Judgments that went into the design of that LLM. Never forget that, you know, even though this is technology, and we can talk a little bit later about how it works, and but you know there has to be human oversight. You know, something as simple as we've all heard about hallucinated cases, yeah, that's pretty easy. You need to have a human oversee it. But you know, anything in our firm that AI touches, We essentially have an audit trail. You know, who did it, what changed, what was added, and somebody has ownership over that. Nothing leaves our law firm without some kind of ownership over what AI has produced. And that's essential. You just cannot let your law firm run roughshod and use AI in any way that they want. But going back to you know Creating the digital twin, so to speak. I had so much information over a career. You know, all these podcasts that I've done, they're all recorded, right? You upload a transcript. You know, all the all the presentations that I've given about you know all different topics. It's something I want to talk about later because congratulations on your book.
Tyson Mutrux (10:40)
Thank you. Appreciate it.
Michael P. McCready (10:41)
I have a book coming. Out at the end of October. So that goes into it. And
Tyson Mutrux (10:47)
Very nice.
Michael P. McCready (10:47)
It's just a matter of giving the LLM enough information to be able to come up with a rational response.
Tyson Mutrux (10:56)
Nice. Speaking of the book, we just hit number one this morning, so that's pretty exciting. We were
Michael P. McCready (10:59)
Awesome.
Tyson Mutrux (11:00)
We were were I was tracking that. It was funny. I didn't care at all and then someone mentioned something. I was like I think I've been like scrolling and checking and refreshing and this morning I checked, it's like number one, so very cool. So get your copy founder optional. Totally did. Absolutely did. But back
Michael P. McCready (11:10)
Gotta get a g s gotta get a screenshot of that. Yeah. Yeah, no, I didn't have a chance to download it and read it, but I'm looking forward to it.
Tyson Mutrux (11:20)
Well let me know what you think. Be brutally honest with me. I'd I can take it. But I wanna get back to you though. I wanna know
Michael P. McCready (11:26)
Mm-hmm.
Tyson Mutrux (11:27)
How broad is the scope of this thing? So i like is it like is it s just case specific or is it can I like ask it like where's the copier paper? Like how broad is it?
Michael P. McCready (11:38)
No, that's a different LLM, okay? So we have another one that has basically firm SOPs. So you remember the old days you had a big binder, right? And you would print off all of your processes. Processes are very important people. You should have them, but you'd have a big binder and you would flip through it all and you know, and then went digital. And we created basically a Wikipedia site to answer questions, right? So all of our SOPs and all the things that we do to run our firm went on to a like a Wikipedia type web you know private just for the firm. And that was that was the logical next step. But now we have all of our SOPs into an LLM. And you know i th the wonderful thing about it is that it will find you the answer without having to look. So for example, we've got an SOP if a return a green card, right? A return receipt comes back undeliverable. I mean we have what we're supposed to do with it, but where the heck do you find it?
Tyson Mutrux (12:46)
Sure.
Michael P. McCready (12:48)
So with our LLM you literally can ask the question, you know, where what do I do when I get a green card? Boom, there's it directs you right to the SOP. But here's the more exciting part about it. Is that the LLM that runs all of our internal processes, including our employee handbook and such, it can come up with an answer even if that specific item is not in there. For example, if you typed in if I if I cursed at my supervisor, would I be fired? Right? That question is not in our handbook. But if you were to put that in, it would say pursuant to section five point four, using vulgarity in the workplace is not acceptable. Pursuant to seven point two, you can be disciplined. Therefore I would conclude that you would You follow me, it would draw that conclusion even though that specific question is not in there. So it's much broader than just having an SOP with answering a question. Going back to the analogy, the Wikipedia page was like Google, right? It directs you to the page that answers your question, but with an LLM, you know, it doesn't need to direct you to a specific page because that page may not exist.
Tyson Mutrux (14:16)
All right, so with this, I'm assuming that there were certain questions that were flowing back up to you, which is why you have the what would what would Michael do. I mean, is that true? I mean that is that assumption correct?
Michael P. McCready (14:29)
You know what? To be to be to be frank, you know, this it's a little bit of a novelty, right? Having this digital twin. I don't want I'm still very much involved in the practice. And anybody can reach out to me for any question that they need, but you know, this is kind of a neat a neat tool. It takes into consideration a lot of my management style and you know how I view my firm, what our culture is, what our values. Are so all of that is baked into it, and people can ask it any I find it much more valuable to ask, you know, not factual questions, right? Is you know, hey, I've been I've got a case manager who's you know continually not documenting things in Smart Advocate. You know, what would Michael recommend my next steps to be? Right? I can guarantee you it's not report me to the boss, that's not how. I handle things, right? So those kind of things. And the other thing is, you know, this could be my legacy, right? This is a way of being able to be available to answer questions without you know without being available. And frankly, the content which we've fed it is so extensive that you know the answers are probably better than what I would give myself. I would tend to be much you know, much shorter, you know, a and this takes into consideration all of that.
Tyson Mutrux (16:01)
I guess what I'm kinda getting at is it did you have this realization where almost like you had become the operating system of the firm where people depended on you a little bit too much? So you're using this as a way offloading some of the
Michael P. McCready (16:16)
Yes, yes, absolutely. You know, my digital twin but also our other our other AI databases and You have to, all right? Is that law firms are going to become technology companies. Plain and simple. We are technology companies, we are service businesses, and our service happens to be law. And that's the future. You know, notice that law becomes secondary. You know, you're never going to replace the you know the decision making and the judgment of a lawyer. But so many other things that go on in a law practice will be enhanced and augmented with the use of technology.
Tyson Mutrux (17:01)
So let's walk through walk me through a little bit about the design of it and how you built it. Like what was the first version like and how did you create that?
Michael P. McCready (17:11)
Sure. We developed you know in an internal LLM system which is you know basically it's just it's called a ChatGPT wrapper. Okay, is it's using the computing power of ChatGPT, but it's limited in what we give it. And I'm simplifying it slightly. But this is not complex. I mean people do this. I know that you're a big fan of hiring people off of Upwork. You can find somebody that can build this out for you. We did it internally, but it's not it's not a heavy lift. The What goes into it, however, you have to have the data. Okay, so AI is not going to make your law firm more efficient if it's already disorganized. And you can't just expect to dump everything in there and expect AI to make sense of it. Now it will it will help, but the more structured your data, and I use the word data very broadly, the more structured the information that you can give. To an LLM, the better your results are going to be. And so, you know, building out the technological framework for an LLM that was fairly easy. And I had, you know, all of this information. You know, I've got SharePoint folders with all my PowerPoints. I had my team go back. I mean, I've got a list of podcasts I've been on, but you know, I went back to all of those and got transcripts. And so, you know, have it, you know, if you have SOPs, standing operating procedures, or an employee handbook, those are both great places to start. So upload those into an LLM and there it there you've started. So with the with what I built out, I just had so much that I was able to give it that it was effective.
Tyson Mutrux (19:12)
So when we build Foxy, we started we started vibe coding with Bolt and then we eventually switched over to vibe coding with Claude Code. And so our CTO Kash of the cool thing is that we have a really good relationship where I come up with these designs, these kind of wild designs, and he actually makes them work. So I'll actually I wonder what was your workflow like? Did you did you all vibe code this? Did you hire an actual or Do you have an in-house coder? And then what
Michael P. McCready (19:42)
Well yeah, we have we have at this point we have in-house. You
Tyson Mutrux (19:46)
Gotcha.
Michael P. McCready (19:46)
Know, with and with any law firm, you know, the first step is that the owner, the manager does everything themselves, right? Then you're in the position to hire people to take some of the things off your plate, you know, whether it's an office manager or an associate or a paralegal. But then there's also things that you absolutely have to outsource. You have to outsource bookkeeping. Right or taxes. A lawyer should not be doing their own bookkeeping. Eventually though you grow and when your firm is a certain size, it makes more sense to bring those things in-house. And we've grown so much over the years that it just made sense to start bringing in, you know, in particular IT in-house. But like you, you know, if It most of them are my ideas. Hey, I've got this crazy idea. And I still do a lot of the stuff. You know, I love to create the instructions for a Claude project. You know, turning it into an agent, you know, that's one step beyond me. And I do let our IT team do that.
Tyson Mutrux (20:53)
Very cool. All right. And you okay, so you'd mentioned the transcripts. You mentioned some other things too that you uploaded into the system or fed into the system. Did you do things like uploading like emails and like things like that where like it could like capture your voice more? Did you do anything like that?
Michael P. McCready (21:10)
No, because I Capturing a voice is very, very important, okay? But with all of my transcripts, you know, it largely catch I mean it is my voice. Now the other thing that I did do is I uploaded deposition transcripts, closing arguments, opening statements that I've done over my career, so it gives it a little bit more of a legal context as well. So we and we have another LLM that the lawyers use more than that, but it also you know can people can use you know my you know WWMMD to get ideas on how I think from a legal process perspective as well.
Tyson Mutrux (21:55)
Have you noticed that has it exposed any inconsistencies in your own thinking? Have you noticed anything like that?
Michael P. McCready (22:02)
No, no. It's remarkably consistent. I mean it's uncanny. One of the one of the presentations that I give is using ChatGPT through trial, okay? And at the very end, I've got a fact pattern. I say, you know, give me a closing argument in the style of Mark Lanier. Okay, he's one of my heroes. I think he's his best trial lawyer of his generation.
Tyson Mutrux (22:22)
Fantastic. Yeah. And a great human being too.
Michael P. McCready (22:25)
Yes,
Tyson Mutrux (22:25)
Yeah.
Michael P. McCready (22:25)
Yes. And so, you know, it gives a transcript and it's vintage mark, right? It's great. But I'm not Mark Lanier. I can't give this closing argument. So I upload three closing arguments that I've given, and I said, give a closing argument in this style. I mean it's unbelievable, right? So what comes out is, I mean, I can stand up there and give that closing argument, you know, in my sleep. It's my it's more than just the words, right? It's also my phrasing. It's it takes into consideration my pauses, my cadence, my you know, it's a transcript, yes, but I can hear myself giving it. So I mean it's just a phenomenal resource.
Tyson Mutrux (23:15)
Okay, so I know that there's probably some people listen to this right now where a lot of their value, their internal value, comes from being the one with all the answers. And I don't I you've never struck me as that guy. You've always been the guy that's been willing to delegate things and kind of push things out outward so that other people can do them. I guess what would your advice be to people that are that are struggling with the idea of sort of them not being the hero anymore and then they're offloading some of this off to an L that can now sort of be the hero, right? Where they can answer those questions. I guess what advice would you give to those people that are hesitant to do that?
Michael P. McCready (23:52)
I've got two things. First is I only know by the title of your book, but probably read your book because the title seems to indicate that. And then the second thing would be read my book, because that's exactly what I talk about. Right is, you know, my the title of my book is and I'm not hawking my book, but it's
Tyson Mutrux (24:12)
No, I want to hear about it, so yeah.
Michael P. McCready (24:13)
Yeah, yeah. It the book is called Built to Run, Designing a Business That Thrives Without You. And you know it's twenty-five plus years of my experience in growing a business. And one of the biggest realizations that you as a an attorney or as a business owner need to come to understand is often you are the pinch point. Okay, you are the one that's holding back. Now, the second thing that I try to emphasize is that everybody has a different definition of success. Right? And so in my book, my book talks about my definition of success has been able to relocate to Puerto Rico. And what steps did I do to be able to design a business that allows me to do that. But you know t for any attorney, you know, look in the mirror and really figure out if you're the pinch point. And I suspect the same thing will happen To most people as to did me is when I had that realization, you know, my whole team, everybody that works for me said, Yeah, Mike, we already knew you were the pinch point. And so it's recognizing that and then getting out of the way. Then, of course, that doesn't mean you know kicking back and drinking a pina colada. It means focusing your time on what is the most valuable use of your time. And the most valuable use of my time for the past five years has not been handling cases, it's been developing and growing the firm, and then also developing and mentoring attorneys and leaders in my firm. So that's what I spent my time on and that's the best use of my time. So technology is a way to leverage all of that.
Tyson Mutrux (26:01)
So and you know this, you kinda mentioned it before where like when you first start, like everything's on you and then as you grow you kind of you transition work to other people, you start to transfer it to other people. What do you think are the biggest mistakes people make whenever they start to transfer that work off of their plate onto someone else's?
Michael P. McCready (26:19)
I think that people misunderstand the word delegate, okay? Delegate does not mean just give it to somebody and forget about it, okay? You've got to delegate, but you still need to make sure that it gets done. It's I don't remember where I read this. I mean you and I both read a lot of books, but it's delegate, don't abdicate. So that's really important. So don't look at it as Don't have to do this anymore. We all have stuff that we don't want to do. And we all have stuff that we should delegate, but that doesn't mean that you have to ignore it. So, you know, making sure that the person that you delegate it to is the right person. Make sure that what you want done is clearly explained, and then some level of accountability. And you will find that if you practice that cadence with delegating, Things get done. And if they don't, if it's not done the way you want, then you need to go and make sure that you explain to the person that's not exactly how I wanted it done. Maybe, maybe not bl I don't blame them. I say I probably didn't explain it properly. Okay, so here's really what I want done, or can you do it this way next time? And this is a really interesting point. But it ties into technology. I hear lawyers all the time. I don't trust AI. You know, it makes mistakes. You know, and AI does make mistakes, but you know what? Humans make mistakes. Many of them. And
Tyson Mutrux (27:50)
Many of them. Many.
Michael P. McCready (27:52)
Here's the difference: if you tell AI you did it wrong, it will not do it wrong again. How many times
Tyson Mutrux (27:59)
Correct.
Michael P. McCready (28:00)
Have you told somebody, a human, you did this wrong, do it this way, and they do it wrong again, they do it wrong again. So, you know, technology is a way of systematizing and making sure that things get done the way you want them to get done, and get done the same way every single time. So you know, so as you're growing a firm, technology to me is just you know another asset, another tool that we have that frees up people's time. You know, if we can use technology to become more efficient in what we do, then you know then we're going to have more bandwidth to do the things that we really should be doing as opposed to the things that can be delegated either to a to a human or to a you know to a to an AI.
Tyson Mutrux (28:52)
So I stole an idea from Brian Mitman. Brian Mitman, he's also just a great guy. He reminds me a lot like you, a lot of you, a br just really smart guy. But he they track in their firm like one of his dashboards, and I'm probably gonna explain this wrong, but it's where like they'll track where things were not done. So the dashboard, a lot of people have the dashboard of like, you know, revenue, leads, all the kind of stuff. And he has all that stuff too, but he has a dashboard where it tracks like fields that have not been filled out by humans, you know. And so what I did was with our case management system, I had it have an AI going behind all of the humans. And okay, all these fields were missed. And then it's it actually will suggest responses for those fields. And it usually gets it right. And so I'm wondering what technology you will are using where you're ch you're making sure that people are getting things right. So where you're talking about the abdication and all that kind of stuff. So
Michael P. McCready (29:48)
Mm-hmm.
Tyson Mutrux (29:49)
Yeah, what technology are you using to make sure people are getting things done and getting it done correctly?
Michael P. McCready (29:53)
Very Similar to what you've said, is that through Claude we will create a project and that project will have instructions, and those instructions will be run the same way every single time. And then through what's called an MCP server, it will connect to our case management system, Smart Advocate, and run these projects as an agent. So that's the that's the conceptual framework. But what we've done is that we've created agents at every step of the process. You know, from intake to opening the case to a 30-day agent, a 90-day agent, a 180-day agent, a medical records agent, demand letter, you know, all the litigation. And what that so I've got a great team, okay? I've got amazing paralegals, amazing case managers, you know, intake. Take settlement paralegals. I've got you know just a superlative group of individuals working on our cases. But they are human and they will miss things on occasion. And AI is there to catch all of that. So you know, off the top of my head, you know, when we open a case, you know, it's going to review the case to see if there's a diminished value claim, right? If it's an auto case. It's going to evaluate whether. Whether there's a body worn camera, BWC, by a police officer. It's going to identify whether there's traffic cameras, whether a preservation letter needs to go out, whether there's indication that perhaps this is a ride share. You know, if we have we have part of that estimates what some of our states are mandatory limits disclosure, some aren't, but we have one that estimates what the policy limits are going to be based on a whole bunch of criteria. So all of this stuff comes you know when the case is opened to a case manager, and there's your roadmap, right? You know, after 30 Days, this is this is probably as closer to what you were talking about. You know, it'll say, hey, we don't have a lane acknowledgement from the defendant yet, or we don't have policy limits, or we haven't you know we haven't gotten the FOIA request back. Also it will review that hey, you know, the medical treatment that the client has is probably sufficient to get the policy limits already. You know, it's analyzing all of these different factors for. The case manager that you know a case manager you know they're very good but they will miss things sometimes and so this is this is just a dedicated you know roadmap throughout the course of the case for my team to follow.
Tyson Mutrux (32:45)
Yeah, so we have a task that pops up right around the three-week mark for attorneys where they're to review the file for any red flags. Because sometimes you don't like at the very beginning, you don't know everything. And even at three weeks, sometimes you don't on a PI case. So we have it go through and w what we're going task by task and seeing if AI can replace the task. And so what we did is we replaced that where we have the AI go through and analyze the file. Look for red flags and it cuts it down the time substantially. So now the attorney goes in, it creates a task for the attorney after it's done it. The attorney reviews the red flags that have been identified by the AI. And then it all then they also review the file too. So they do both. So you got a human checking on things. But it is a massive time saver and it catches so many things like what you're talking about. It catches so many things that the humans will normally miss. And so because I mean like especially attorneys we're like we're super busy. So we're always like, okay, I've got I've got court, and then I gotta come back from court. I got I got a client phone call, now I've got to draft this pleading, all these different things. So it just it streamlines things, but it also identifies things because we're always we're always in a hurry. So it's extrem I mean it's extremely helpful. I think it's cool the way you're using it too. I do wonder, because are you're still using Smart Advocate, right? Okay, so
Michael P. McCready (34:01)
Mm-hmm. Yes.
Tyson Mutrux (34:02)
So your so your MCP it's able to connect with Smart Advocate in Claude. Is that how it works? Excellent. So are you
Michael P. McCready (34:09)
Correct, correct.
Tyson Mutrux (34:12)
Is your team do you have I guess do you have a Claude team plan that's able to the so that everyone has access to Claude? Is that how you're doing it?
Michael P. McCready (34:19)
With the MCP, it's not as because it runs on agents, we don't need as many licenses as you know as for the number of people that we have. Originally before the MCP, yes. In order to run a Claude project, you needed a license. And truthfully, you know, the licenses are not particularly expensive as long as you stay within your token usage. So, you know, Two years ago everybody was talking about hallucinations, you know, now everybody's talking about blowing through, you know Wads of money by token usage. And maybe if you're not familiar with it, is that every time you use AI, it costs a token, So the bigger the project, the more information that AI needs to digest, the more tokens that you're going to use. You get a certain number of tokens with each monthly subscription. If you choose to go over that, then you spend more. You know my controller and I are the only ones who have control over approving asking for more tokens. Everybody just has to learn to work within the token limit. And for most of what we do, that's very easily done. You know we're not we're not running, you know, huge, you know, physics computational I mean What lawyers do from a from a language standpoint and from a data standpoint is not a particularly heavy lift for AI. So we've been able to live within the licenses that we have without necessarily blowing the budget.
Tyson Mutrux (35:54)
Yeah, so like for day to day work, I it's a very low risk of actually running through a bl blowing through a budget. Like we have blown through the budget the last three weeks, but because we're in the middle of three massive builds and so it's just that's why and like when you're building you can t 'cause Kashev and I are both building at the same time. So it's a lot of we're burning through quite a bit of tokens. But that's a rare like once everything is I mean everything's stable right now, but once we've gotten through these builds and the builds are stable, then it's not gonna be a problem. We will like this week I don't think we'll actually burn through it. So we'll this will be the first week in three that we won't, so not a big deal.
Michael P. McCready (36:28)
You know, it just occurred to me, Tyson, listen, you and I both know what we're talking about, and our firms are fairly advanced when it comes to this, but let's just step back, okay, for a law firm that you know really hasn't done anything with AI. Okay, I've got a subscription to ChatGPT and I kind of know how to use it and I'll go back and forth. But you know, here is a very easy thing for you to do. Okay? Now An advocate of ChatGPT in law school, there was Westlaw and Lexus, whichever one you were more comfortable with, they were largely the same. I felt that, except that Claude has definitely moved ahead with regards to legal side. So, you know, I tell people, you know, go with Claude. If you're gonna go with one, and we do we actually go with multiple, go with Claude. And what we do is after ever we analyze what our intake team's decisions are on cases. Okay, this is very easy to do. If you're doing your own intakes, then you know I guess it's not as relevant. But what we do is you know, we look to see what the facts are and what the decision was by the intake person. And we do a couple things. One is we rate the call and how good the intake person was with the client and suggest feedback. But more importantly, you know, we are looking to make sure that. Made the right decision. Not just, you know, hey, we declined the case and we should have accepted it, but also we accepted it and we shouldn't have. So that's a really, really easy first project that you can come up with, is you know, and not all of us are PI, but you know, you should have an idea of you know what qualifies for a potential case or a client and what doesn't. And you can use criteria to allow AI to go back and decide whether the decision maker was correct in accepting or declining the case. If they decline the case and they shouldn't have, get on the phone, right? Get on the phone and get back to that client. So that's a really easy project to do. And I would encourage anybody who's you know skeptical or hasn't gone down that path, that might be the first place to start. And one Once you see, and guess what? Claude will help you do this. Right?
Tyson Mutrux (38:53)
Mm-hmm.
Michael P. McCready (38:54)
Right? Ask Claude. I am a such-and-such attorney. I want to build a project that determines whether we are accepting or declining the proper cases. What are my next steps? There you go. Boom, boom, boom, boom, boom. It's gonna ask you. Well, I'm gonna need your screening criteria. Okay, you upload the screening criteria. Okay. What's the next step you want me to follow? Then what instructions should you should you draft for me? And then you patch the instructions into the project. I mean it really is that easy to walk down this path to develop a fully functioning you know Claude project.
Tyson Mutrux (39:37)
I really like the idea of identifying that client that you're looking for. Cause it's easy to do if you just think about your best client and create a profile for your best client and think about your worst client, create a profile for your worst client. That'll the worst client will give you the red flags. Because we all know what a basic case is. Like if you're a criminal defense attorney, you know you're not taking wills and estates, right? You're not taking those that stuff. So that stuff's easy. The harder ones are the okay, like what are the ones we're looking for? And then which are the cases that we that sounds like a case, but we want to avoid these cause these red flags. So just think about best client, worst client, and you can create profiles around that, feed those into claw and do everything that you just said. So I think it's really it is simpler than what people think. I think people think of these LLMs as these really complicated things. Yeah, on the back end they're super complicated and technical. But for the things that we need them to do, it's very simple. You just tell it to do it. And
Michael P. McCready (40:30)
Nice.
Tyson Mutrux (40:31)
If you don't if it here's the thing, if you don't know how to do it, ask it. Say hey I here's the idea I have. I'm not really sure where to start. Can you tell me where to start? And it will tell you where to start. That's what's cool about it. Is it's self-educating, which is also pretty cool. But
Michael P. McCready (40:48)
Yeah, I think that I think in the past that lawyers you know, especially of a certain generation, could afford to allow someone else in your firm to do the technology. I think those days are past. This is a tool that every lawyer absolutely must learn how to use. Do you need to do all the crazy stuff that Tyson and I are doing? No, okay, but you need to understand how an LLM works and start using it for your practice. And The more that you use it, the better it will become. And then once again, ask it questions. Right? I'm a you know, I'm a bankruptcy lawyer. How can you help me? Right? Or, you know, how what are 10 things that you can do that will help my paralegal complete tasks such and such? I mean, it really is that easy. In and that's the place that you need, every lawyer needs to start.
Tyson Mutrux (41:42)
So I'm trying to look up that. I can't remember if I don't know if you know the name of this company. So David Haskins, I really like the idea of what you're talking about with the leads because you just made me think of something. David Haskins had created a platform because he's with Drive Law Firm Marketing. So if I can't pull it up and remember the name of it, I'll just reach out to David. But it's speed. Ai, where I don't know if you're aware that
Michael P. McCready (42:03)
Sure. Yep.
Tyson Mutrux (42:04)
Yeah, so it'll analyze your leads team and the intakes. And it finds cases for you, which I think is really interesting where it will go through and you can do the same thing with your own, but it you made me think of that whenever you were talking about what you're using it for. Sounds like that's kind of what you're doing. Like you're having analyze that.
Michael P. McCready (42:19)
It's you know, listen not everything that we do technology wise was my idea, okay? There are a lot of things that there's a lot of tech law tech companies out there who've got amazing products, but you know, the technology behind them is not very difficult. And these things can, you know, be largely brought in-house. And the and the cost To do so is going down and down and down. So, you know, when were building stuff out a couple years ago, it was you know it was pretty expensive, it's just becoming less and less expensive. This is stuff that you know, solos and two and three lawyer law firms can have access to this technology themselves, or you know, you can you can hire a vendor. And speed AI is excellent, it's a it's a it's got you know a great idea. A lot of these law technology. Companies stem from an idea. Here's a problem. How can we fix it with technology? And you know, and there's definitely value to that, but at a certain level, you know, these are things that can be can be brought in-house.
Tyson Mutrux (43:32)
All right, so want transition a little bit and I want to ask about the move because as you said, you're in Puerto Rico right now. And I just want to know what was the cause you I don't think we've stressed this enough. And I'm gonna brag on you for a second, Michael. Like Michael has a very big, large firm that is a very it's I don't know how many employees you're up to now, but you are you have a very well established firm and When I heard that you had moved, I was very surprised because I've always thought of you as a Chicago guy and like because you've I've been watching you slowly spread your tentacles across the Midwest. And you've done I've always I've always complimented you, your business acumen. I think you're a fantastic person, fantastic human, a great
Michael P. McCready (44:17)
Keep it
Tyson Mutrux (44:17)
Business leader. So I was very surprised when you'd moved. So I what was the conversation like with your team when you said, Hey, I'm moving?
Michael P. McCready (44:26)
That's a great question because I don't think anybody's asked me what that was. So, first, I didn't build my firm with the expectation to move to Puerto Rico. Why did I move to Puerto Rico? First is there are very, very good tax advantages for somebody in my position. I mean, I'll be perfectly frank. I'm not the first one to do this. There's a lot of mainland lawyers that live here in Puerto Rico. So the tax advantages, if you're interested, look up Act 60. I'll tell you. Second is my wife was getting a little bit tired of the winters in Chicago. Third, we love to travel and you know are considering traveling around the world, like I know some of your other guests have done. But she said we need a home base. And so we discovered Puerto Rico and all the benefits that it had. Know and I would encourage anybody, you know, this is this is this can be done. However, you need to build your firm to be able to do this. This is not a matter of you move to Puerto Rico and all of a sudden you have to try to continue running your firm. So I was here for about six months before you know it was generally known around my firm that I had done this. Wasn't hiding it, but I didn't need it to be an announcement. We now have 180 employees and 24 lawyers in seven states. So I wasn't physically in those offices very much anyway. We had built out this firm to be a multiple office firm before COVID. So were doing Zooms before COVID, were doing electronic signatures before COVID, all of this stuff, and when COVID hit. It you know, that was just a blip for us. I sent everybody home with their voiceover IP phones on Friday and Monday were up and running. We never missed a P. But you know, moving here. I wanted to make sure that well, first as I have a great team, I'm still very involved. I find myself working more here than I did back home. But I'm working from paradise. So about a year in Maybe 14, 15 months in, we do a firm pulse survey periodically, right? How are things going? How are you feeling? How can we improve? So forth and so on. And I asked at the time, I want a question is Michael's relocation to Puerto Rico is five very positive, four positive, three neutral, two somewhat negative, one very negative. And we had about 120. Responses. Two people said very negative. And one of them was an attorney who quit two weeks later and said, ever
Tyson Mutrux (47:12)
Sure.
Michael P. McCready (47:12)
Since Michael moved to Puerto Rico. So I don't know who
Tyson Mutrux (47:15)
Yeah.
Michael P. McCready (47:15)
The other one was. But you know and then there were 19 that said slightly negative. A hundred out of a hundred and twenty people were positive or neutral. So, you know, it tells me that what I did is working. And you know it reinforces the idea that this can be done. Once again, this is this is delegate, not abdicate. I'm still very involved in the practice you know, do doing podcasts or speaking at AAJ or NTL, you know, those are the things that I need to be doing and increasing the my profile and the profile of the firm because we're primarily referral-based. So you know, it's an interesting question. You know, it was just sort of, you know, it wasn't a secret, but it just sort of got out. And by the time there was an announcement, everybody was like, and well, okay. So that's how it went down.
Tyson Mutrux (48:16)
I love it. Michael McCready, thank you so much. Really I really appreciate you doing this and sharing your knowledge.
Michael P. McCready (48:21)
Enjoyed catching up with you Tyson.
Tyson Mutrux (48:23)
We'll have you back on whenever the book launches too. Make sure people mention the book again. Is it on for if it's on for pre sale, let us know. Okay.
Michael P. McCready (48:28)
Yeah, it's on it's on pre-order or pre-sale on, you know, Amazon and Simon and Schuster, all that other stuff. It's called Built to Run, Designing a Business That Thrives Without You. And a lot of the stuff that I talked about is all in the book. It's how I've built my firm, right? It's the mistakes, it's the things I learned. It's about my digital twin, right? This is this is my way of getting, you know. What I have done out to other people and I'm hopeful that it resonates with some people.
Tyson Mutrux (49:03)
I what I'll do is I'll get a copy and then I'll whenever it comes out I'll give a I'll give away a copy on the podcast. So I'll make sure everyone can get a copy. Thanks, Michael. Appreciate it. You too. See you
Michael P. McCready (49:08)
Of course, of course. Appreciate it. All right, have a good one. Bye.
Tyson Mutrux (49:13)
Thanks Everybody.