The Future of Legal Teams: Lean, Tech-Driven and Creative

Speaker A: I wish people could agree in advance on how many bargaining points in total they had and then rank their issues in order of agreement. And then, you know, out of that you'd spit, I think the right negotiated contract.

Speaker B: Welcome to Modern Business Operations where we

Speaker A: talk with leaders about how OPS is adapting to our modern world.

Speaker B: Hello everyone. Welcome to another episode of Modern Business Operations. My name is Agui. I'm the CEO and founder of Tonkin. And today I have the great pleasure of hosting Mark Mendel. Mark, thank you for joining.

Speaker A: Thank you so much for having me.

Speaker B: Well, Mark is the SVP and general counsel of the coaching company Exos, which I'm curious to hear more about. Uh, but you also served as general counsel in several other companies. Big, uh, and some of them very known. So maybe let's jump in and kind of hear more about your background and uh, how you got to do what you're doing. Sure.

Speaker A: I had a couple of jobs before I ever went to law school. Those were in technology. And that was a thread that carried through to me post law school. So I spent about nine years as a corporate and securities lawyer working at various firms, but always with uh, a side of technology maintained there. And then I had an opportunity to become the first general counsel of an ad tech company. And the founders there picked me from the pile, I think both because I had this broad corporate experience, but also because I was a practitioner myself and sort of understood the plumbing of ad tech maybe better than others would. And so I've, I'm a three time general counsel now. So there was a, the ad tech company, a smart home devices company, and now a coaching company that is heavily leveraging technology. Um, so I can tell you a little bit more about Exos, uh, if you like.

Speaker B: Well, I think, you know, especially for anyone who likes sports, it's pretty interesting space. So yeah, maybe, maybe in a few

Speaker A: words, yeah, Exos is an interesting company. Coaching can mean a lot of different things depending on what level you're at in meeting us. So we work with professional athletes, we have a big NFL combine training program, we work with Olympians, some elite military, all of that we're helping people to perform at their very, very best. Measured, you know, times in milliseconds and have a, bring a lot of the latest equipment, high tech gear in order to help measure and manage the best outcomes. We roll that into a excess methodology that we then deploy at usually large international or American corporations for their employees as part of a coaching program. And so a lot of the coaching we do is helping everyday People perform at their best in work and in life, but it's inspired by what we do at the high end and what we've learned through that process.

Speaker B: The most obvious follow up question for me then is do you guys use it internally?

Speaker A: Oh, absolutely. It's a hybrid model. So we have in person services at fitness facilities that are owned by Exos and then at our client sites and then there's also an app and the app is for a hybrid workforce so you can meet your exos coach inside the app or in person. And we uh, take good advantage of our own methodology and services internally. It's highly encouraged.

Speaker B: That's actually super interesting. Uh, I loved when there's those cross reference, cross fields when you bring one methodology, one technology from one area into another. And obviously anyone that has this competitive mindset and high achieving always, we always look at Olympians, uh, and you know, you know, NFL as an example of like the top league of any sport as a way to trying to learn from it. What are the, you know, few things as you, as you come in and you kind of compare it to your previous roles. Obviously, uh, uh, gc, general consult, managing group, sometimes, uh, quite big, you know how, how like what are the things you took already from it and applied?

Speaker A: Yeah, so it's, it's a holistic sort of application. Right. It centers around movement, mindset, nutrition, recovery are the sort of four tenets of it. And we're encouraged to do everything from taking a recharge break in the middle of a long meeting to stacking our work week into four longer days and having Friday recovery. So there's a lot of practices that I found to be, you know, just really helpful to coming, coming into work with the right mindset and, and being productive. Uh, and it's nice, it's nice to be able to, to do that ourselves. The other thing that's interesting for us is, you know, the legal department is relatively small. We were, I think, you know, five or six people at our largest size. And um, so it's important to really show up your best when you have such a small team. And so for me that's been a combination of technology and efficiency hacks and then putting the, you know, the Exos methodology to work.

Speaker B: So let's move to that for a second. Uh, you know, uh, you mentioned you started, you were the first GC or started the kind of the legal department and the first ad company you mentioned. Then here being able to serve a larger group with a small team and have efficiency, maybe share a little Bit like what, what, what are the few things you're, you guys are doing that you felt really, uh, moved the needle.

Speaker A: Yeah. At exos in particular, I was not the first general counsel and that was a very different experience from June Group, the ad tech where I was. And there was more of a blank slate. At exos, I walked into a well functioning department that wasn't functioning at its peak. And part of the reason for that was there was a technology layer, but it was not being exploited the way that it could be at its best. And I'll give you a great example, which was at some point there was one document management system that was more like a data room than a true clm. They transitioned that to a popular CLM product. And when I opened and looked what was under the hood, half of the agreements were in a folder that said imported from other system. And they were just there, not categorized. Um, I thought, how are we going to search over this and try to develop trends? We ended up having to fully reboot that in addition to a lot of other processes. And this was all three, four years ago, before AI. And then once we got access to the large language models, we've just really been on a very upward trajectory in terms of tech adoption and success.

Speaker B: This is obviously extremely popular topic right now, but I think a lot of people are talking about it with opinions and it sounds like you guys already are leveraging, uh, on the ground. Do you mind sharing some more specific example where you guys found success? Because obviously a lot of people, especially in departments like legal, are looking for ways that are practical and kind of go around the hype after the hype. Okay, we got the hype, we got the excitement. Give me the goodies. Like, what are the actual examples?

Speaker A: Yeah, I'm going to build up to a couple actual examples, but I want to, um, set a little bit of the prelude, which is we had an opportunity to bring on board someone to the legal department who was to serve as a legal operations professional rather than an administrative support function, which had been the role historically. So that person has a background in product management and technology and a little bit even in the legal space, although I don't think that was necessary. And so he and I have had this great collaboration as to technical minds, with him really being able to focus a lot of his time on these tasks and bouncing things off of me to get the right trajectory. And so that's been a huge unlock. If there's not someone whose primary mission it is to sort of elevate things, it's really hard to make the time for it. We've explored several of the more legal focused LLM tools and then also the general purpose tools for us. With the budget we have, we have gotten sort of more value per dollar for sure out of the ChatGPTs and the Geminis and the, you know, the more general purpose frontier models. And that's what we've really gravitated to. We know they're improving the fastest and they get access to the latest features right away. And so a couple examples how we put them to use. We do a lot of commercial contract work at exos, as anybody in the service business does. We get client MSAs that have insurance requirements in them. What's interesting about figuring out whether or not you can comply with an insurance requirement is every bit of information that's needed to answer any question about our insurance programs is within the four corners of a policy document or five. And so having that complete knowledge base, you can put that small set of documents into a custom GPT with some background instructions and just copy and paste in from a contract and say, do I comply? And it will go through and check off, uh, the limits and the notification periods. And so giving the AI very specific tasks like that has been very successful. Sometimes there are judgment calls to be made and rather than relying on AI as your yes or no approval, it will give people, um, a nudge in the right direction. So another, I'll call it sort of long form analytical prompt, but I put together, uh, just a few weeks ago. We've had a lot of requests to evaluate AI tools in the organization. They all have terms of service. There are only a, you know, a handful of points that we really care about. Are they training on our data? Do they have confidentiality obligations? Do we own all the outputs? You could imagine, uh, what, what's there?

Speaker B: Right.

Speaker A: And so we couldn't leave it to the AI to say, yes, this is approved or not approved. But what we did was we gave the prompt to all of our field staff and corporate staff who are interested in using new tools and said, run this first. It may tell you these terms, I think they're awful, they're not likely to succeed here. They're pretty good, you know, in either case, what it's going to say is now please take the output of this prompt and send it to legal for review. That has stopped people from submitting terrible terms of service along with a thank you note that they weren't wasting their time or mine. And given the lawyers, uh, leg up on a quick review because the details are sort of pulled out into an appendix from the prompt's output, and it's really made that particular task much more efficient.

Speaker B: Well, I honestly want to applaud to you. I think, I think you're, you're really describing the ideal way of handling, in my opinion as a technologist, the ideal way of, of, of businesses to, uh, look at new technology and, and, and look for the leverage in it, uh, in both of the points you made. So the first point of understanding that you need an op, you know, an operation person that is a product person, a technical person that sits within your team and uh, and, and becomes your sort of like deputy for, for, for that and for. And for the leverage that it can bring, as well as, you know, the, the depth of which you guys seem to kind of like understand where the, where these LLMs are, um, good at, where they're not as good at, where there's people good at, where these people are not good at and kind of what they appreciate. So that, that is, uh, that is fascinating and I think very easy for people to also listen to this and, and, and, and understand. Wait, we can do, we can do something like that too.

Speaker A: A lot of, A lot of homegrown resources that have been very useful.

Speaker B: Yeah. And honestly, that's kind of what was my next question was how does the rest of the organization, how does the attorneys. How do they kind of feel about this? It is obviously very exciting. As a leader, you can definitely, I'm sure, see the results. But how does the rest of the organization culturally reacting to this?

Speaker A: Sure. So like, my management philosophy, right. Is based on Daniel Pink, if you or the audience is familiar with them. But it's all about giving people mastery, autonomy and purpose to take to their work. And so everybody has been excited to see how AI can help them in, you know, master a new craft, succeed at it. It's such a great teacher. It's just been inspiring. I mean, for, for myself, I've gotten way down into like publishing a couple of tools and, you know, things that I made that took me longer to make than I could have to do it. But I was just inspired and I'm seeing more sort of projects. Right. We have a Made with AI Slack channel, and people are throwing what they're doing in there and people are one. Upping each other. So, um, it's a super exciting time to be building something.

Speaker B: This episode is brought to you by Tonkin. Tonkin is the operating system for business operations, providing businesses with the building blocks to orchestrate Any process with no code or change management required. Contact us@tonkin.com to learn how you can build complex processes fast. That's another great, uh, practical tip, um, that made by AI. I agree with you. I think we see it a lot of places. The, the hobbyists, you know, might be the first one to pick it up. But if they get recognition and people kind of see the value of it, a lot of other people might say, wait, that doesn't look too hard. I want to, I want to try it out myself. We think the, if we go from practical for a second to a little bit philosophical, how much do you think this is? This is a step change versus this is something that would, in, uh, I don't know, a few years would become a new standard, a new baseline.

Speaker A: I mean I'm, I'm, I'm very much in the revolutionary camp. Similar from the, you know, the transition to having business not computerized to business being computerized only instead of 25 or 30 years, depending on what industry you're looking at, I think it's going to happen in five. And it's, it's just the adoption is just so fun to watch and the improvement in just the last couple of years. I'm, I'm not a sort of doomsday person when it comes to job elimination or any of the potential downsides. I think, I think just the, you know, the, the upside is clearly going to, going to be there as, as it was for, you know, PCs and, and the rest of the big technology that we've, we've seen influence business.

Speaker B: I'm with you. I, I agree. How do you think it will, or maybe even go more specific, you know, if a college, uh, student, you know, studying law right now, come to you for like, advice, what is going to change from that perspective? How you going to hire, you know, attorneys or how would you, the next generation of them giving like the one that are going to the workforce when AI is already established there? How's that look? Because I don't think jobs are going to go away. I agree, but they are going to change. We both agree on that. What is your sort of like immediate prediction and then, or immediate um, reaction and then what is your predictions, you know, going forward?

Speaker A: I think it's going to put people who are more creative in charge because the people who can see the potential and how to get the best results out of AI are going to win the race, so to speak. And so even though law may not be traditionally thought of as a creative discipline, actually really Good lawyering has always been creative and crafting an argument or finding a solution. But it's going to become so much more important when you language models are leveling the playing field and the prompts and how you use the tool is going to be the differentiator. And so that's, I actually look forward to that sort of outside of the legal realm. I think about art and what that's going to become and like the creativity that will unlock that I can't even imagine, you know, sort of what's, what's coming. But yeah, I would, I would say be prepared to try to think differently because that'll get you differentiated output and I think people will need to really differentiate themselves. And the other thing is relationships will always continue to be important whether it's among team members or between a vendor and a client. That's not going away. And so that human connection may be even more influential than ever if vendors capabilities, you know, come to be more and more similar because it makes it easy to introduce a new feature and attain feature parity with, you know, someone else. So yeah, I would, I would double down on relationship building as well.

Speaker B: I don't know if a million dollar. But the dollar question that I think comes up when you think about law and legal compared to other industries that are being, you know, adopting and disrupt getting disrupted some way with AI, obviously, you know, a lot of the things we talked about is very similar to what you know, engineering would, you know, feel and marketing would feel and attorneys specifically in legal firms are slightly different because they're still working on that idea of hourly uh, rate. And I'm curious your thought of from a business perspective, how is that going to change because is that model can survive. What's going on with AI, do you think?

Speaker A: It's a great question. I think that a lot of law firms have been very slowly, um, gravitating away from the billable hour. And this may be an accelerator, but there are so many unknowns in the facts when you begin a new legal matter that that's always going to be difficult. But the AI may be able to play referee to easily come up with a um, change order. Right sort of thing you might think of in an engineering or a construction project that is not present in legal invoices today. Right. I estimated something, I went over my estimate, you know, sorry, let's have a discussion about that. But here's, here's why it'll be transparent to everybody. And so I think the AI will bring a really important level of fairness and so how the price was derived, if it was hourly or if it was project based, I think will be, uh, a mix depending on what the task is. You'll be paid more for the real value you bring. So similar to the how well can you prompt will equal how great is the output. And I think people will see the benefits of that and we'll continue to pay for good advice, but at a rate that seems fair. And certainly some of the costs are coming down overall. The hourly rates are going up, of course, but the time spent should go down as well.

Speaker B: Yeah, I'm curious to see how fast this will happen if you compare it to other, uh, disciplines that has sort of went through a similar path. By that I mean you used to pay per hour for uh, engineering work, you know, software work and construction work. You would used to pay per hour for like design work, you know. But you know, bunch of those have transformed completely into being like delivery based and project based or sometimes even just I don't know what the, I don't know what the cost would be. Like, we'll do it first and kind of like, you know, then we'd be able to estimate like how do you estimate art? You know, it's like not, it's definitely on the time that he took to draw the damn thing. Right. So it's like you have like a factor to it that is greater than the, than the effort. And when technology comes in, any technology, it really changes drastically the definition of work from what is the kind of manual effort versus like what is the. Your point? Creativity or expertise or taste that the service provider brings to the table. Right. Uh, and craft. Right. Mastery. So like.

Speaker A: Yeah. Ah, so I'll give you another example, right. Of sort of where I think things are going. I think a lot of people are already just asking AI for their contract instead of asking their lawyer for the contract. I just hear this in entrepreneurial circles constantly that what I got was good enough, we'll see where it goes. And then I see myself doing it too. Right. Here's a specialty matter that I could send to an IP expert. I'm more of a generalist, but let me have a chat about it. And you know what, I feel comfortable now and I'm maybe not going to outsource this bit of work. I'm going to DIY it. And so I think, you know, outside law firms there'll probably be some, some shift further shift towards more in house teams and less reliance on outside counsel. But I think there will also be an acceptance of more Slop in contracts. People spend all this time negotiating the perfect contract only to put it in the drawer. Hopefully it's a bad day when you have to read your contract after you've negotiated and signed it. It's only when something has gone wrong. And so we all know AI makes mistakes and that'll never be perfect. It's getting better. But I think people will say this is so fast and good enough is good enough that the acceptable number of mistakes and agreements is going to become higher. Uh, in part because the efficiency outweighs the time that you would have otherwise spent and in part because AI will end up being a fair adjudicator on the back end. Right. And um, I'm waiting for the first sort of AI arbitration clause or um, there are people working on it already. But for it to become more commonplace because it's just going to speed everything up and uh, you know, allow people to take more risks.

Speaker B: You know, if we go even further into the, into that realm, which I think very interesting your takes on this. I had a conversation, you know, a few weeks ago, uh, on a parallel example, but that's I think apply here. If there's an AI that is creating an RFP and then an AI that is filling the RFP and then an AI that is reviewing the rfp, do we need an rfp? You know what I mean? Like what is the point of an RFP at that point? Maybe there's a. Because obviously you need that communication. But maybe the version of the way we think about RFP right now is for us human to try to consume the information. There might be a better way for. If the AI does it on both sides, maybe there's. Right. So this is an RFP example where I think it's more interesting on contract unlike RFP is that at the end of the day contracts are a way between you and me to, to agree on things. And like you said for the worst, we agreeing ahead of time on the worst case scenario in which actually none of us will judge. A judge will judge. Now what I've heard and kind of my experience and uh, peripheral experience we know with, with folks I know many times you have negotiate stuff in a contract. You talked about commercial contracts. So I'm taking that as an example and. But you go to a judge and a judge can still decide while this is saying that there is this issue or that issue that is not contractual but more fair on that way or another and therefore, you know, decide differently. So if that's already the case where a contract is not so like set in stone. To your point, how does that even like um, it's just super interesting to me. Like is that a week on a path where some of those fundamental concepts can be completely rethought about. So I can, I'm going far, right?

Speaker A: Yeah, no, no, and I'm, and I'm m gonna, I'm gonna go right there with you.

Speaker B: Right.

Speaker A: Like we have case law and case law is part of what makes it difficult to go to court and have a predictable answer. And the judge is obviously variable and you know, to some extent you ask the same LLM, the same question twice, you'll also, there's variability there, but you could imagine codifying 5,000 examples, right, because there's no scarcity of data of everything that could go wrong. And you and I can't think of everything that could go wrong, but AI could just come up with every example, right? And instead of having a human readable contract, what if it was more like a compiled software where all these sort of pre agreed dispute resolution mechanisms or outcomes were baked into the contract. And yeah, there was like a, you know, term sheet that you and I could read on top of it. But really if you wanted to ask, if you wanted to ask a question about the contract, your default would be just chatting with the LLM because that's where all the detail lived rather than page flipping a long document with a, uh, you know, an expert attorney there. So I've never thought about that sort of use case where the contract itself becomes primarily machine readable. But you know, now that you mention

Speaker B: it, I think it's, it's even further than that. Like there's something, there's something fundamentally flawed in my personal opinion in the way, you know, a lot of those things. And it is actually what I'm about to say is similar to the RFE scenario and the contract scenario and, and there's, there's other examples too where you know, if you compare it to something else, like let's say a um, service order you mentioned earlier or like, you know, things that are like more transactional where you're like, hey, this is like a term sheet you mentioned. Like this is what I'm, this is what I'm buying, this is what I'm selling. Are we agreeing that that's, that that's the project we're trying to do together. And ah, that's true for like employment as well. Like I'm hiring you for this job. Is this the job you are? You know, and that's for, for that price. Are we agreeing on that? Great. Like that term sheet, like the, the, the executive summary for that matter is like that is the, that is actually the real contract socially. It's like this is what I've kind of like all the fine print. The problem with it, it's not even, it's about dispute. There's a little bit of uh, this is my personal opinion. Okay. No, I'm not representing any, any company, anything. It's my, you know, it's not necessarily in good faith. And what I mean by that is everyone, I just want to do it a little bit of a land grab of overprotecting themselves from someone from the other side that the other side might be completely. Okay. But in the case that they're trying to do the same, then I'm gonna, you know, kind of protect myself. Right. You see it with insurance, you know, there's like, you have a lot of different, like things that might be hurtful actually for if you have a visitor in your restaurant or if you have an employee or if you know, whatever, because they're trying to protect from the 0.1% of insurance fraudsters that are gonna try to take advantage of it. Right. So like all in, all in. All in, all in, all in all you have justice system to uh, basically assess the specific case to your point and say, well that none of it matters because this what happens here and there. So it's really interesting to me what happens when we can cut through the kind of bs, go completely and say like it actually doesn't matter what you write here because people are going to try to hack it. It's like a little bit like the security thing too, like the locksmith and the thief and they keep running in circles improving the lock just to improve the hack and so on. So I'm fascinated by how sometimes technology can make uh, more than just an incremental shift. It can completely help us rewrite things that might actually not benefiting the two parties. Both parties want to do something together and they just want to agree on what it is that they're doing and not sort of like, uh, you know, sign their life away or whatnot. Yeah.

Speaker A: So two reactions to that. One is, I think this plays right into the idea of self adjudicating contracts. Right. Which is the name I think we'll call the thing that I was describing earlier where the, the real information about the party's agreement is encoded rather than in English. And then the other thing is this would all be so much easier if parties could decide on their relative bargaining power.

Speaker B: Right.

Speaker A: Because people are going to stick it to people because they can, because they want to. It's part of risk management. Right. Stick it to somebody is a little bit unfair of a term, perhaps, but it's, it's really about risk management. And if you're the vendor, you have to take all the responsibility in the world for your tool and you're going to say no. If there's a hacker that's an external party that's come and showed up here and this is your business, and if, you know some third party impacts your business and I happen to be the vector, that's not my responsibility. And these things go around and around and around. I wish people could, um, agree in advance on how many bargaining points in total they had and then rank their issues in order of agreement. And then, you know, out of that you'd spit, I think, the right negotiated contract.

Speaker B: Yeah, I like your point about the 5,000, you know, edge cases, but I also kind of like the 5000 examples is also a matter because this is a great example. We can kind of finish on that point. But just the example you just gave with, you know, the hacker use a vendor as a vessel and all that stuff. What I found interesting is what matters at the end is not even your bargain chip is what the, is what the judge will decide. You know, I mean, like at the end, at the end of the day, there's going to be a complex scenario that happens if this happens ever. And in that complex scenario, there's going to be complicated evidence that happens this or that. And at the end of the day, it almost doesn't matter what the contract says. It matters like if everyone acted in good faith. It's one thing if, if there was anyone that did not act in good faith. It doesn't matter what the contract says, you know, and if, and if there's a situation where one side would lose nothing from this, while the other side will be completely dead by, by it, and the judge will say this is not a fair result even. You know what I mean? Like there's, uh, a, there's an interest. I just find it almost crazy that, you know, that the power at the end of the day of the judiciary system is both significant as well as, though not as coded as some, and some people expect. And, and how do we can use, to your point now, with something that is no longer a strain of knowledge, you can use it for a vast amount of variation and knowledge to maybe reduce the point of, of bargaining chips and more into like this is almost doesn't matter because 99.99999999 of the times, if this happens, then it doesn't matter. And maybe it will allow, to your point, the focus to be on the right things.

Speaker A: The soft factors are always in play, right? In the example of the security incident, the big issue is going to be did the vendor use reasonable security measures? And that just opens up a huge can of evidence that needs to be brought in. And those are. But that's all language. That's the sort of thing that the LLMs would be good at parsing. So I'm interested in what happens to choice of law.

Speaker B: Right.

Speaker A: Like right now, if you want someone to interpret your contract as literally as possible, you know, go ahead and pick New York law for a commercial agreement. Um, they're just going to give you what you said, even if it's a little bit unfair. Other jurisdictions have a tendency to be, you know, weigh some more of the fairness factors that you mentioned. Maybe one state, you know, could be Nevada has been, you know, innovating a little bit. Maybe they'll adopt an interesting sort of choice of law that is basically entirely on an AI first judicial system.

Speaker B: Interesting. Fascinating, really. Well, that was great. I think we talked a lot, covered a lot of things. One thing I always like to ask on a personal note, we started your career and we also touched on new attorneys, uh, or new law grads. What is maybe an advice you received early in your career that he took with you and worth paying forward?

Speaker A: I think the best advice I ever received wasn't like a pithy sentence. But at ah, the first law firm that I worked for, Thatcher Profit and Wood, there was a partner who took me under his wing. And what he did was every time I was going to send an email to a client, he said, send it to me first. And he would rewrite it and rewrite it and rewrite it until, you know, after nine or 12 months he was like, no comments, no comments, no comments. And so once I had learned to write the way he wanted me to write, then I was at last free to not have to run it by him. And that was such great instruction that the way I pay that forward is often the people on my team have the substantive issues, right, but the way you package things is really important. And so I'm like, hey, show me the COVID note. And in that way try to get people to, to sell what they're, what they're attaching to the document or attaching to the email.

Speaker B: Nice packaging, packaging, packaging. Mark, thank you so much. I learned a lot. Uh, this was, uh, very insightful. If anyone wants to continue to nerd with you on those things or pick your brain, where can they find you?

Speaker A: LinkedIn is my only public platform. Uh, Mark, D E L Exos. Um, Exos. Put those into Google. My profile should come right up.

Speaker B: Awesome. Thank you again. Thank you so much.

Speaker A: You got it. Thank you. Sonny.

Speaker C: I hope you enjoyed this episode of Modern Business Operations. You can see the show notes and all of the resources mentioned on Today's episode@tonkin.com mbopod thank you for listening, and be sure to subscribe for updates on future future episodes. And if you're interested in staying up to date on the latest in business operations technology, head over to tonkin dot com.