Building Agents & Removing Friction, with Kelly McCord of SimpliPhi

Speaker A: It's really cool what these agents can do, even simple ones. There's also very complex things that you can be building, but it's just about trying. And I will say my experience with the building agents started rocking. The videos are very cool in what you can do, but it takes trial and error. It's about, uh, doing the reps. There's a lot of learning and iterating, but that feedback loop of like, when you get it to do the thing that you were trying to get it to do, it's magic. That's fantastic.

Speaker B: I'm Hal Marcus.

Speaker C: And I'm Mei Maangwudiwe.

Speaker B: And this is Meeting of the Minds, the legal AI podcast from evisort.

Speaker C: We interview lawyers, professors, and legal operations pioneers that are pushing the envelope leaders

Speaker B: using technology to drive great business outcomes and shape the future of the profession. Make may may. We are about to share a really interesting conversation we had just the, uh, other day with Kelly McCord.

Speaker C: Yeah, no, I'm really excited to share this with folks. I've met Kelly. We've hung out at a few different legalops.com events and dinners, and she's really quite the dynamo, you know, her career includes significant tenure at hsbc where she served as the head of legal operations, though she's currently the founder and managing director of her own consulting firm, Simplify Management, llc.

Speaker B: Yeah, I really enjoy talking to Kelly for a number of reasons, and one of them is that her career path has very little to do with legal, until suddenly it did. She started in engineering, she went through business analysis, and then found herself in this very important role with legal and Legal Ops, where she obviously achieved a lot. We talked about her simultaneously using, like, literally simultaneously using and benchmarking multiple LLMs against each other, building on top of the results of one with another, which is something I've not heard qu quite that way before. She talked also a lot about the practical realities of where you should move fast, where you should move slow. Sometimes you need to do both of those at the same time.

Speaker C: Yeah, she's really a pioneer and so well steeped in these different kinds of AIs and LLMs. I think it's going to be a great episode. Let's kick it off. Kelly McCord, welcome to Meeting of the Minds.

Speaker A: Thanks for having me. I'm excited to be here.

Speaker C: Would love for folks to hear a little bit about your history and career path and how, uh, you got to where you are today at Simplify.

Speaker A: Yeah, absolutely. It was not intentional, that's for sure. I was always really Good at math and science in school. And so I wanted to be an engineer. I wanted to build airplanes. I thought that would be fun. And I started going to job fairs and decided, no, no, thank you. That doesn't sound great at all. So my professional career really started at PwC doing their business process consulting. So essentially all of the auditing that happens before you get to the financial statements and then from there, someone who I had worked with at PwC was tapping on the shoulder and said, like, hey, do you know anyone with your background who might want to come join this hedge fund? We need someone to be our business analyst. We're building a function. And I was like, well, I could do that. So I joined Fortress and led a couple of their IT teams for their credit funds, really business analysis and credit and customer support. So got to do that, build out a couple really neat tech solutions for these financial services stakeholders. And that also probably the most fun job I ever had and took years off of my life because they will work you very, very hard if you can. Or if they can rather. So about the time my body needed a break, someone else who I had worked with previously again kind of tapped me on the shoulder and said, hey, we're over here at HSBC. You remember when I left PwC before and I needed a work life balance. This place over here has a great culture, great work life balance, but we have these things called consent orders, which I guess aren't great. And you're pretty good at solving problems. Why don't you come in and help us build a solution for getting out of consent orders? So I joined the regulatory liaison team and at the time, that group fell under the legal department. And just the timing being what it was, this is right around the onset of legal should also be run like a business. Our CFOs of companies are now also asking legal departments to look at their balance sheets and run numbers. So our general counsel was like, oh, I need a coo. I need someone who knows how to run a business. And one of his directs who was in my management chain was like, Kelly, did you look at this job description? Are you interested? And I'm like, I don't know what a COO is, but we talked about it and it is essentially understanding how processes work, stopping doing dumb things, trying to start, do a little bit more smart things and be good with numbers. I'm like, okay, I think I can do that. So fast forward. I ended up getting the chief of staff to general counsel job coo, then renamed ahead of legal ops but ultimately, each step along the way it's been trying to solve problems, trying to understand how complex systems and people work, and then just making life easier. So simplifying, right? While I was at hsbc, I first had to learn what Legal Ops was. And then we kind of redesigned the program through a multi year transformation. And at the culmination of all that, we were able to combine our legal operations in North America, from US and uh, latinum into an Americas. And at that time I stepped away and let someone else manage it. And you know what? I think I'm gonna go out on my own because it's really fun doing this, like going in, helping corporate legal departments to solve problems. So I started consulting. At first I thought it would be a temporary thing and now I've come to really love it. But when I was doing that transition, I had some soul searching because I started auditing it. Regulatory, legal, like, who do I wanna work with? Even though the nature of what I'm doing is a lot of the same and, and turns out I love working with lawyers. I think they're great and they think I'm great. It's a fantastic relationship.

Speaker B: You know why they think you're great? You're the only one that's ever said, I love working with lawyers.

Speaker A: I think that might be true. People think it's crazy. But when you work with front office people and you work with the tech and you work with the folks who are more on the forefront of technology and then you deliver a solution that you're happy with, you know, the feedback is often cool, why don't we have that yesterday? And of course you were going to do that. But when I work with lawyers who. The business of law, the foundation of law is there's a lot of precedent, there's a lot of looking back in history and you have to appreciate that, right? To be great at what you do, you have to understand where we've come from. But they're not all as forward looking, or at least you're not trained to be. So when I deliver a solution that I'm proud of, that I think is pretty neat, a lot of them think it's black magic. And so the feedback that I get is just incredibly positive. And so I want to keep delivering like it's a symbiotic relationship. I love it. So when I was choosing to form my business, simplify, it's a play on words because Phi is Greek. Simple. Anyway, you can look it up. What I was looking to for my business, I wanted to stay Working with lawyers, I think it's fun. And not just lawyers, but the legal profession, legal ops. I've really found an ecosystem, a community of people who want to raise the bar together. There's not a lot of competition that I see. There's no backstabbing, there's is. Everyone just wants to share what we know, help each other. And just what is that? They say the rising tide raises all boats or something along those. That's what I get from the legal ops community and I'm just really happy to be part of it.

Speaker B: Yeah. Because lawyers, never competitive, never antagonistic, they're never trying to win the argument, all those things.

Speaker A: Maybe I just get to deal with the best. I don't know.

Speaker B: Yeah, I'm only half serious with this one because I am a lawyer. I've stayed in the profession though, through many, many years of focus on technology, but largely because I also really like working with lawyers. But I mean, I'm well attuned to all of the criticisms that come in from anyone in tech, anyone in business, anyone in sales that suddenly finds themselves working with legal. And they just go, oh my God, the lawyers. Oh my God, the expectations, the detail, you know, because a lot of us come up through the law firm environment where it's, you know, fiefdoms and manage it yourself and forget mentorship. And a lot of these things really are true and have been true for a long time and they're getting a bit better, but they're ingrained in the mentality. Is that a factor for you as you work with counsel? Do you work more with like in house counsel law firms?

Speaker A: Yeah, so typically in house counsel, and I think one of the things that I'm going to call it a superpower is I'm not great at, ah, reading between the lines or picking up on subtleties. So if I am engaging with someone who might have like a little bit of extra snark and it's like these subtle like, nudges, it's probably just going over my head. So where others might be offended, I'm just completely oblivious. Like I'm only getting what's like really coming through. And so I'll take that as a win. But also when I was going to school for engineering, like, I'm a rocket scientist. Okay. And like not fully fledged rocket scientist, but people like to say it. And I went to school for rocket science, so it counts. And so like the whole ego thing that sometimes comes with dealing with attorneys who have gone through really, really tough schooling. Yeah, I did too. Like we both have our battle scars. It's okay. So I think that part of that I don't worry too much about on the personality side. And then as far as high expectations. Yeah, me too. I tend to be rather precise. So if I'm going to be overly precise, working with attorneys who are overly precise. Okay, that's great. I love it. So I think that there's definitely, uh, some alignment in the personality, the traits that are common in lawyers. That kind of works well for me. But, yeah, I mean, absolutely. I've worked with a couple who are like, ah, you know, I've done X, Y, Z. I'm clearly better. And I'm like, okay, that's fine. Your colleague over here values what me and my team can do, so I'm gonna help serve them first. I'm not gonna force what I can do on you. And then when the FOMO kicks in. Because that is a common trait.

Speaker B: Yes.

Speaker A: When the FOMO kicks in and they're like, wait, but I want what he or she has. And I'll be like, okay, I'm happy to engage in a conversation now that you want what I've got. Okay, let's go. So it works out. I think I'm also lucky to be far enough in my career. I've been doing this for over 20 years that I'm no longer trying to impress anyone, and that was definitely a factor earlier in my career. I spent a lot of my twenties doing the things I thought I was supposed to to do, and now I just do the things that bring me joy and help people, and it works out.

Speaker B: That's fantastic. It sounds like you've really plugged into how do you turn that competitive spirit and that adversarial kind of mentality that gets drilled in in school, if it wasn't already there, and turn it into a positive. It's like, this is how you keep bettering your practices.

Speaker A: I will match energy. I will try to match energy. So if you're positive, I'm right there with you. If you are not, then maybe I don't want to engage, and that's okay.

Speaker C: There might be a lot of folks listening who are inspired by your ability to say, hey, I want to go out on my own, hang up a shingle, start an organization, do this at scale for folks who might be on the fence thinking, hey, I've done this long enough that I think I'm good enough to do it anywhere. But they might have that fear factor about kind of taking that more entrepreneurial route. Do you have any Words of wisdom or advice for anyone in that situation.

Speaker A: Yes, I've got the logical, practical and the emotional. So my gut response is, do it. If you're feeling it, do it. You'll be happier than ever. My logical response is like, also just double check your bank account to make sure that you've got that buffer because there's that uncertainty, especially when you're starting off. I went through a bit of an emotional roller coaster in the first six months. It was like, I've never, uh, marketed before. I don't know how to market myself, I don't know how to do sales. I know how to do what I do and I think I do it well. But there was that unknown uncertainty fear of, well, how am I actually going to get projects? And that was probably not even six months. I think it's about three months. But it was at the very beginning and you don't know how long that's gonna go on. So there was that fear for me of like, how do I get my first client? What do I need to do? As soon as I started talking to my network, just telling people this is what I'm interested in. I haven't had to go look for clients. People come to me. So that fear of like, who am I gonna do work for if I'm working for me, it just, it works out like there's so much need for what we do. So much need for what we do. And especially now with what I'm seeing this year, every corporate legal department is stretched thin from budgeting perspectives. Like there's going to be internal pressures to get a FTE approved. Definitely easier to get an approval for a part time consultant or contractor. So you're in a good situation from the cfo, budgeting people who are, uh, a constraint on your potential clients. And if you've got the skill set, just do it.

Speaker B: So what are the kinds of projects that you get pulled into or at least the ones you're really excited about when they happen?

Speaker A: That's not always the same question because sometimes I get pulled into projects that I'm not excited about and I don't know how to say no. I just really like being able to solve problems and make someone's life better. Like if there's friction in a process and I'm very process oriented, so it's okay. What is causing you agita? What is that pebble in your shoe that if it doesn't go away is going to be a blister and then going to be crippling to you? So where can we remove friction? And to do that I've begun understanding the process. I love a good process flow. You can't see it, but over here I've got a giant post it board because I'm constantly like drawing. So understanding where the flow is and then removing the pieces of friction and, and that as a concept can be applied to multiple different types of projects. Right. It can be contract process mapping or parts of a contract process. So it could be document management system, it could be legal intake, it could be outside, uh, council management. And so it's just who has a problem and once it's solved and then I just dive in and do what I can.

Speaker B: Are the solutions you're working on always technical solutions? Do they tend to be more process combined with, with tech? Like other than the sticky notes, how do you piece this apart and tackle it?

Speaker A: If I had my way, it would always involve tech. I think if there's something that is repeatable, if you can describe to me what a process is and I can write it down with a series of if then statements, because ultimately that's what a lot of it comes down to, is just a, uh, complicated series of if then statements. Right. I'm oversimplifying. Then we can probably automate the majority of these repeatable routine things and then free up time for actually using your brain for judgment and analysis. There's so many things where we spend a lot of time prepping the material to be able to then do where we add value. So an example in the contracting world, like just to create the initial draft of a contract, right? Like, let's hope we've got a template. Right, let's hope we've got a template and a rough playbook. But to populate that template, you need certain bits of information. Where is that information coming from? Well, it's probably not coming from you directly, it's maybe coming from sales, maybe coming from procurement. But when you're getting it, are you getting it in an email that you then have to copy and paste or another human has to copy and paste and then make a couple edits? Or maybe you've gotten third party paper and it comes in and you're trying to manually compare it to your template. So you have to do this bit of reconciliation or creation before you can shift into value add mode, where you're now looking at the deliverable and saying, okay, does this make sense? What do I need to tweak? Is this protecting us from risk? Right, so all that before stuff, if we can articulate what is the process that we go through to get to where we can now start providing value. Not to say that copying and pasting isn't adding value because someone's gotta do it, it's gotta get done.

Speaker B: I scored really well on that in law school. I just wanna say copy and paste. I aced it.

Speaker A: Well, I'm sorry, my friend. That is a task that's probably going away in the future.

Speaker B: Really? Damn.

Speaker A: At least to some extent. Or reconciliation between systems. Right. I was talking to someone who' working with their E billing tool and accounts payable. Right. Because we've all worked with them. You've got an E billing tool, invoices come in and then you have to actually send it to another system to have the payment made. And then ideally you would like the source system, um, to be notified that the payment was made so that people know. Well, you got to reconcile these two systems. How much time is wasted reconciling the data just to figure out was it paid, when was it paid? So we can give it. These are the same activities that we're doing over and over again. Let's just make it easier. And these are like the simple stuff. This is before you even add in gen AI capabilities into what's possible. Like, don't get me started. So I think it starts with understanding the process, understanding the people's pain points, trying to meet people with where they are. I do try to leverage whatever tools are already in house. So if your team has already been approved to use xyz, what can we do quickly with what's already been approved? And then if there's something that would level up the way that you're operating, then we can talk about different providers that are out there that can do so much more. But sometimes first, especially for corporate legal departments because of budgetary requirements, sometimes you need to deliver something like the baby version of a solution to whet the appetite of those who are opening up the purse. I think it really, it helps to be able to see a baby solution first to get people to want, oh, but can I do this and that and the other also it's like, well, no, it can't. But any of these five tools could. Would you like to get demos? So process. But a lot of tech, I love that.

Speaker C: And I mean, you said not to get you started when it comes to generative AI, but I will like to get you started on that. I think we were at a dinner@legalops.com a few months ago and I was really struck by your wide breadth of knowledge across the gen landscape from chatgpt to Gemini to Perplexity. I think even right before the call we were talking about a recent update to Excel's AI and such. And so really thought it'd be cool just for folks to kind of almost tap into a mind that is so attuned to these different kind of tools and capabilities. But what are some differences between different LLMs that you think might be good for folks to know? Like, hey, I use ChatGPT in these situations, Gemini in these situations, Perplexity in these situations. Just for someone who's kind of new to this space to get some grounding level, just any kind of knowledge you might be able to impart to them,

Speaker A: I will do that. And I'm pulling on my phone to show you. This is the page for my work related stuff and I've got in row perplexity, Gemini, Claude, ChatGPT manuscript, which are all LLMs. And what's missing here is Utari, which is a new LLM that is currently in development and I'm one of the beta testers. My AI coach created it and it's super cool, you guys, but it's not out for public. So I'm using that one on the desktop, not on my phone. So I've got these five, four of which I use quite frequently and sometimes in tandem with each other. I just recently upgraded my office setup and in doing so I was having three concurrent chats with Gemini, Claude and OpenAI or ChatGPT rather. I started with the same prompt and then I would go between the different tabs and see what was suggested and then continue that conversation, go to the next one, what was suggested, continue that conversation. And this went on. And then I would take a feedback from this one and pop it into that and say, hey, what do you think about that?

Speaker B: Are you doing the multiple LLMs at once thing to compare and contrast? Or are you leveraging what you learned from one, using it for another? Like, is there some force multiplier you're getting?

Speaker A: Yeah, absolutely. Because I'll get a response from one that the others did not consider and I like it. And so I'll feed that into the others to be like, hey, what do you think about this too? And so I'm able to build on top of each other and then it comes together, right? It's like how you say one plus one equals three. This is one plus one plus one is, I don't know, 15. It's great.

Speaker B: I'm thinking we should call this cross training. I'm going to Trademark that.

Speaker A: Okay, I'm all for it.

Speaker B: Your thing. We'll go into business together on this.

Speaker A: So I can tell you what I use them for and it's not always a one for one. I will still use the vernacular I got to Google that. But I don't actually Google things. I Perplexity things. Perplexity has for the last two plus years been my replacement for Google. I like it, I'm a fan. So Perplexity is my go to for quick information, stuff like that. ChatGPT is my go to for a lot of personal things, a lot of like planning things like my nutrition, working out, supplements, planning routines. Part of that is just because it was the first one that I really developed good conversations with. So if you go into the settings and the models learn, the more you interact with them. So in the settings I've got listed already things that I want it to remember about me and about my history and so it can lean on that as we're building through and iterating. I also like ChatGPT because of their custom GPTs. It's where I first started playing more with the custom GPTs and I've built a couple solutions for clients of mine that are specific to them and I think OpenAI ChatGPT is the most widely known and like generally mainstream if you will, for friends and family also. I don't know if you guys had seen this, but a couple months ago OpenAI opened up a group chat feature which I really like. So I use it again more for a personal thing. So when we did our year end goal setting, my partner and I had a group Chat going with ChatGPT to talk about our goals and both of us were able to contribute to the conversation and we were co iterating on um, what our annual goals would be. So I like the group chat feature there, but I use it more for personal type things.

Speaker B: The clients you're working with can't possibly be doing the kind of experimentation that you're doing. They can't possibly make such a wide use of the generic models. Do you find that they are like locked into one generic model and just sort of building gems, building their own specific versions or are they not even really there for the most part and they're turning to you to turn these generic models into something that's more of an application for them?

Speaker A: That's a great question. Most of my clients are the larger institutions and so I am seeing that they tend to be locked into one or two bigger LLMs. Everyone that I work with is Also a Microsoft shop. So they've got Copilot. They might only have Copilot Chat or they might have copilot for, um, M365 I think is what it's called. So with the agentic tools and I have it for my personal business as well. But usually people have like ChatGPT, Enterprise or Perplexity. But how widely that's embedded throughout the organization I think varies. Everyone wants to be using the AI. Like we got this AI for you guys, go ahead and use it. And then there will be a number of training sessions or genius bars or whatever you have, where people come and they get to see examples of how you can use the AI. But a lot of it, uh, in my observation, is these institutions basically giving a tool to the users and saying, do the things, you know, the things. And most people are using it for their writing. I use it for writing. Like, again, I put different LLMs against each other to make my writing better. It's great. And so I think a lot of people are using it for emails, but not a ton of folks are creating agents, creating custom GPTs or gems, creating tools with these resources that they've been given. I think a lot of it is learning. I think a lot of folks, if given a pre, uh, configured tool, would gladly use it. I'm not sure that asking the users to build their own tools is going to be the most fruitful approach. And so where I come in, because I love doing it, is I'll have a conversation with people and it's like, oh, gee, wouldn't this be great? Or I'll just have an idea of like, wouldn't this be great? And then I just spin something up and give it to people. So last year I was helping with a DMS implementation and at the very early stages of it, we spun up a DMS WE Governance and Design Task Force. So it was a group of 15ish people from across the legal department who needed to make certain decisions around the design of the tool. Right. The security model, the different infrastructure, et cetera. We recorded all of the meetings, we had the PDFs or the PowerPoints from all the meetings, we had the transcripts. And even though my memory is okay, I thought, gee, wouldn't it be nice if we loaded all of this knowledge into a custom GPT for the project team and then we can just query, hey, what was it that the task force decided? Or can you just review all of these transcripts and synthesize our, uh, recommendations for the Sterico? Wouldn't that be great. And it did, and it was great. And we shared it with the project team. And even though I'm no longer that client, they have this knowledge that was over a year ago and the tools available today to do something similar to that, actually you can do things so much better, so much more advanced. But that was just one idea is like, well, we have this knowledge, we have this data. Why aren't we making it more accessible? Right. And so that was just one example of the thing that got spun up. I'm coming in to help build the solution once I understand what is it that we're working with. Right. We want to start with where people are. Even though Gen AI is super cool and the capabilities are very robust, human behavior is slow to change. So even though we're all very excited about it, I think we have to pace ourselves in an institutional environment. In my personal life, I am going warp speed. Like I'm building things left and right. I'm having so much fun. But on the client side, trying to pace myself completely.

Speaker C: Understand, you have so many skills beyond just drafting, prompt engineering and things like that. When we think about the future of the skills necessary for folks in the Legal Ops space, do you think the level of familiarity that you have with these AI tools, do you think that's what more and more Legal Ops folks are going to need to have? Or is there still space for folks to not have that level? I'm kind of curious where you see that going.

Speaker A: I don't think it's necessary to be as intense as I have found myself to be. I've gone down a bit of a rabbit hole. I like it here. It's fun, but I don't think it's necessary. I think that I have value, so I'm not mad about it for my own self and where I can help integrate within teams in building solutions or helping to translate into plain English, like what a solution is doing so people can understand. I like being able to connect those dots. Folks will need to have at least a basic understanding of how the machines are built, how they're working together, what an LLM is. You don't need to know how to build the tools. That's okay. But you have to understand kind of what's happening so that you don't just take whatever you're getting and believe that that's true and then you go act on it and then you're mad because that wasn't right.

Speaker C: Do you have tips for folks who might want to, uh, up level in this space?

Speaker A: Just dedicated two hours a week, one hour a week to reading the news or watching videos, and one hour to practicing. If you can do that, you will get better. You will build the skillset. Every single one of us coming up in the industry, we did the reps. We are good at whatever it is that we do because we did the reps. This is the same as anything else. You just got to put in the time to practice.

Speaker C: Kelly, in this agentic world we're living in, uh, where folks are all talking about building agents, you and the few people I know really out there actually building agents. And I would love to have you just talk a little bit about what that experience actually looks like. Even in this kind of nascent era of agents, what you kind of see it doing, maybe we see it maybe not doing for a long time, maybe never doing. Maybe it's things it's already doing now that folks should be aware of. I just feel like we'd be remiss not to give you a little bit of platform to talk about that while it's such a hot topic.

Speaker A: Topic, yeah, absolutely. Agents are super cool. Guys I have built for my own personal work. I have not built any for corporate clients, if we're going by the strict definition of agents. But it's really neat to think that you can give a set of instructions and you can get a set of capabilities and tools to an agent, a bot, an assistant, a technical assistant made of ones and zeros, and send it off on its way with a purpose, and it runs. And it's also within your control how far it'll run. So a lot of times these agents, you can either give it full control, which some people do. I do not. I'm going fast, but I'm going slow. Right? But you can say, like, just go do this thing. I saw a video recently where someone had told its agent had sent a picture of, uh, I forget what it was, but just call it like a knickknack lying around the house. Sent a picture of it to the agent saying, I need to sell this in the next month for at least $25. And then, uh, gave the agent access to its Facebook account, Facebook marketplace, a couple other things, email. And it went off and run. Like it went and sent messages and set a distance and set timing and pricing and marketing, like, sales, like it created the post itself, and ultimately it sold. And I thought, that's really cool. I wonder if it's just a video. And so I tried and it did it. I didn't actually sell it Because I had the guardrails on where I was like, you don't have the authority to actually agree to the transaction, but you can do the other things. And it would ask me along the way, like, am I, I created this post, Am I allowed to post it? Yes, go for it. So it's really cool. I think first people need to get comfortable with the idea of automating. And then once you have a couple of things that are automated, you just pop a little agent on top and that agent will run the automation for you and talk to the other agent who's running the automation over here and then pop it out to you. Maybe only when a threshold is meet, like, okay, keep on doing this until something is within this threshold triggers and then let me know. Otherwise, send me a weekly report just so I know what you did. Like, it's really cool what these agents can do, even simple ones. There's also very complex things that you can be building, but it's just about trying. And I will say my experience with the building agents started rocky. The videos are very cool and what you can do, but it takes trial and error. It's about doing the reps. There's a lot of learning and iterating. But that feedback loop of like when you get it to do the thing that you were trying to get it to do, it's magic, it's fantastic.

Speaker B: As you're talking about that, my mind is going back to all the discussion eight, nine years ago around smart contracts. As we saw at the time, blockchain looking like it was going to be more of a foundation for a lot more agreements than has potentially happened other than with very specific cryptocurrency kinds of transactions. And one of the big concerns there was these things are self executing, self operating. So where's the human control? And one of the key answers was you can put breakers in, you can put these points in where no, somebody's gotta flip the switch. And bless that. I think we're seeing the same kind of discussions and dynamics play out around agents. It's not a yes or no, it's not a they have complete and total autonomy or they don't. There are ways to set these things up with appropriate levels of autonomy.

Speaker A: A hundred percent. I see a day where the majority of contract negotiation is done between the agents. Right. Because they've already got again, you have to have your ducks in order. You have to have some of those foundational layers. You need to actually have a playbook or tolerances. Right. And it doesn't have to be like a formal playbook. But you need to know what is your risk tolerance, what are your, uh, approved fallback provisions? Right. In general, directional guidance. But once you've got that, you plug an agent into it with the right tools. It's drafting the initial contract or ingesting the third party contract, comparing it against the parameters, kicking back some red lines and kicking them back to the email address of your counterparty. Their agent is picking it up out of that email, doing their own additional red lines based off of their own set of thresholds and risk tolerances, emailing it back to our email box, which is then picked up by our agent. And maybe there's a counter so it could go back in numeral times, or maybe there's a counter of like once it's gone through three turns. Now pop it to a human with a brief summary at the top of the email. Because we're never going to get away from email. Lawyers love email. So pop it in an email with a summary to the attorney of like, hey, here's what we've already done. Here's what we see as risk. Here's your attachment. You do that final check and then we'll kick it over. That's very doable. And if you have a system like that where on both sides of the conversation you've got agents doing that review, imagine how much time is reduced because you already know what your tolerances are. I mean, generally speaking, like, you know what you're going to be comfortable with on an initial or secondary pass, of course. And then you get the human in the loop for those exceptions where it's like, how big is this relationship do we really need? Or is it so small? Like, why are we bothering even talking about this? Right? So that's where you get the human for those exceptions. But if you got agents on both sides within an hour, you've got the almost fully negotiated contract. Like, almost. Oh my goodness, it's doable. And if enough people buy in, like, it's not that far away.

Speaker B: Yeah. So much more business can be done when it's less about the back and forth on the artful language to capture the terms, and it's the back and forth instead on the terms because you know where you're agreeing and where you're not. Although as many of our guests have pointed out, you don't have that documented a lot of the time. So that's getting your house in order and your data, like you were saying before, hey, we've covered a lot of ground here. I do Want to ask one last thing because I think your particular background is really relevant for this. You didn't come to Legal without going through a couple of other departments first, if you know what I mean. You've seen a lot of different aspects of the business and then found yourself in this kind of important role in legal. So with that in mind, how can Legal, with the benefit of technology, AI, automation, forward thinking, what have you better integrate and collaborate with the rest of the business?

Speaker A: I am a big fan of uh, starting with the business objectives. So it has been a challenge with every legal department I've talked to to concisely articulate the value that a legal function provides to the broader company. What I like to do is start with the business objectives. So if they are already stating our goal is to grow X percent here or develop this segment or increase this, that or the other, and this is where getting to know your counterparts in your lines of business. So if there's a, uh, chief of staff or chief operating officer in that line of business who's tracking their quarterly updates to exco or whatever, get from them, what is it that we're tracking? So if we find out, what is it that they're scoring themselves on, trace back to how are we as a legal department enabling that and have the conversation in their vernacular.

Speaker C: Right.

Speaker A: They don't necessarily care about the minutiae of what it takes to get it done. But we can say you wanted to increase this metric. Well, we handled this many contracts or our turnaround time was reduced by this to help facilitate you closing more of those deals. You are welcome. Thank you for being a great partner. Can I please have headcount for a legal Ops person? But really putting it back in their terminology to help show the value is great. But also, and this is all relationship driven, having someone on those stereo within the lines of business, not a legal operations person, but from the legal department, not just being a partner to their senior leaders, but being on their management team through a dotted line or what have you sitting in at the onset as you're talking about a new transaction or a new product line or a new development so that you can be adding in some of the thoughts around the risk and the way that we're framing a new product before it gets to the point of oh hey legal, we want to do this thing, can you make it happen? So getting yourselves embedded in there is definitely helpful in promoting the value of the legal department. And I think as you shift minds about how the legal department is perceived, you get engaged in more things. And earlier on, legal and finance and compliance, we all sit in a unique area where we see all of what's happening in the businesses. Each line of business is generally very, very aware of what's happening in their world and maybe generally aware of what's happening elsewhere. But in legal, we get to see it all and we can help connect those dots. And so the more that we're brought into the conversation and we show that value, the more that we're going to keep getting invited back.

Speaker B: Once they learn it's not the principal's office, they're a little happier to be in the room.

Speaker A: Yes, exactly.

Speaker C: Well, thank you so much for your insights today, Kelly. It's been a pleasure chatting with you.

Speaker A: Yeah, thanks for having me. This has been fun.

Speaker C: Meeting of the Minds the Legal AI Podcast is brought to you by Eversort. To learn more about Eversort and how we can help you contract better with AI, visit evisort evisor t.com um, you

Speaker B: can find Meaning of the Minds on Apple Podcasts, Spotify, or wherever else you listen to podcasts. Don't forget to click subscribe so you won't miss future episodes.

Speaker C: On behalf of everyone here at evisort, thanks for tuning in. M.