Name: Can your AI do this? The fastest path to value.
Uploaded: 2026-09-16T12:46:22.783Z
Duration: 1 h 1 min 18 s
Description: Can your AI do this? The fastest path to value.

Transcript for "Can your AI do this? The fastest path to value.": Thank you. Thank you, Jen. And audience, you know, I you know how I love Well, some of you do if you've been to our other sessions, but, I love being the dumbest person in the room, and that that is the case today because, you're in for a treat. I think Stefan's Sunni's knowledge on contract technology, legal ops, AI far outstrips my own, and I promise no one is going to say that aside from right now, no one's gonna say the phrase, AI is not gonna replace lawyers, just lawyers that don't use AI. So, we're gonna hopefully focus on some fun things that are interesting for you and, share some materials as well. I know, we had a question around slides, already. And so at the end, you know, you can you'll get an email with all these materials, and we have some QR codes as well for you to, you know, download materials and stuff like that. So we're here to talk about AI. But before Steph, Sumi, and I get started, let's go to our first poll for this session. So tell us a bit about what your biggest pain point with contracts is today. You know, lack of visibility, some exotic problem that's unique to yours. If if you choose other, please, if you're willing, share and chat what your unique pain point is with contracts because I think we'd love to hear from you. Steph, Subhi, you know, looking at these poll questions while the audience kinda starts, you know, answering this, What do you typically see in this space as far as contracts go and pay points? Yeah. The it's a new thing. Yeah. I I I'm looking at I'm like, oh, kind of all of them. I no. But I think overall, contracts and noncontracts, it's the too much manual work. You know, I think there's been some fits and starts in the industry around tree, you know, contract terms and obligations management and and and people have somehow navigated those and been able to come up with some temporary solutions, or workarounds. But it's the manual work. I think that people are really the redlining and the tracking and the reporting, all of that is still kind of, I think, at the top of people's minds. Alright. Yeah. And view. Well, you stole my line. I was gonna start by saying AI won't replace lawyers. No kidding. But, but I I think Sumi is exactly right. When people come to us and they say they wanna automate contracts, the first question Sumi asks always is, well, what part? Because that's really what kinda gets to the heart of it. But it really is around all aspects of the contract life cycle management. Sometimes they want a smart repository. You know? Sometimes they want the redlining. So it it's really all over the map. So I'm very curious how people end up responding to the service. Well, let's let's let's find out. Alright. Too much manual works. So so Zoom call them. Got my finger on the pulse. Yeah. There you go. It's it's and I mean, you know, some of these overlap. Right? It's it's just it's too slow, but it's because of too much manual work, and stuff like that. So I I don't see let me see if I can scroll down. Oh, nobody chose others. So, you know, everyone clearly knows, that they fall into a very similar buckets when it comes to contracting. So that's interesting. And we're gonna talk, obviously, about how AI can help solve some of these pain points as well. So why don't we just jump straight in, Steph and Sue, instead of kind of, you know, defining what AI is and stuff like that? Let's just talk about the problem. And I wonder if you see this with your clients as well. Because when I was at Clark, what seems like just yesterday, but it's really weeks ago now, everyone was saying the same thing. We're getting pressure to use AI for legal more, you know, top down, be more efficient, you know, whether it's because of constrained resources or people just want to be more efficient. Is that what you're seeing as well when you talk to legal teams? A 100%. I've never seen this before, and I'm as you know now, but I've been doing this a really long time. And, you know, in all my years of of running legal operations, I've never had a CEO come to the general counsel and say, hey. You need to implement e billing or whatever. That just never happens. And, and so now this is the first time where legal is really feeling pressure to implement AI. We're seeing it all over the place. So people are calling us saying, hey. We need an AI strategy. We need to do something around AI. We're getting a lot of pressure from the top. You know, it mostly GCs want to do it. They wanna start exploring. So we're seeing so much less resistance, I think. But they are definitely getting pressure. But re very recently, like, within the last couple of days, a survey came out showing that CFOs aren't quite yet seeing the return on investment. And so I'm curious to see if that pressure continues. Yeah. I I I will have to see how it plays out, but for sure. Yeah. I think it's it's interesting because unlike some of maybe the other initiatives, AI is one of those that legally has to get involved with because they're up upfront, the company is asking, well, what can we do? And so, I think the onus is now on the legal teams to actually explore, understand, and then come up with a policy to, or at least they should be coming up with a policy. You know, they're a stakeholder for sure. There are others that should be involved, but, there there are real legal risks, that need to be considered. And so I think people, this is one where where they're really pushing legal teams to to start exploring and using it. Yes. I chuckle I chuckle a little bit whenever I see this happening just because, you know, when you're in legal and legal ops step, like, what you would say, you know, it's it's an uphill battle to get approvals for technology. You know, you have the Yeah. 15 business cases, attend a 100 meetings, jump through a bunch of hoops, and that's not to say there's no hoops for AI. Right? Because that these exist. Two, sometimes the hoops are on fire as well. To just get to the point where you can start doing vendor selection for, you know, an e billing tool or matter management or a CLM tool. And and now it's just, hey. Go find some AI. Let's start using it. You know? But It went sorry, Navin. I was just gonna say it's so funny. And if you if you ever talk to Brandy on our team who's our our head of innovation, it it's so funny. It went from, like, no. Either there's no AI at all or maybe we're allowed to use Copilot, not even, Copilot Studio, just Copilot, to now clients are the same clients are saying, now we want agents. And now we're before, we were like, well, try it. Try it. Now we're like, woah. Slow your roll a little bit. Like, you really don't want agents running loose on your machine. You really don't. You know, that needs to be managed and locked down, and it's just kind of funny to see, you're you're you're talking about, you know, how the this change in mindset happens so fast. It's just it's really interesting to watch. It it must be nice for adoption, though, I I suspect. You know, I think part of why, you know, AI really took off I mean, it started originally with chat GTT. It was just, you know, the user experience is just so simple that, you know, you can start using it without training. Of course, we all know you need training to get kind of consistent, accurate, good results out of it and data and all of that. But I think from an adoption standpoint, at least one thing that's been nice about it is it doesn't feel like it's as much of an uphill battle, with the end users to get them to use these AI tools. I wonder if, you agree or disagree with that sentiment. Yeah. I I think that is is the underlying key challenge in figuring out what the ROI is. You'll hear Brandy kinda talk about this a lot is that there are so many use cases, but unless you're capturing the information around it, in a systematic way, how do you really, identify what the return on investment is? Yeah. You know? And we've we've tried to do that, and we'll talk a little bit more, later on about how we have done that with our clients. But you really do have to, one, experiment to understand, and then two, start to, you know, prioritize what are the ones that you're gonna focus on. Mhmm. But it does take some discipline. It's it's, I think before we move on from this slide, kinda how I summarize this. The more things change, the more they stay the same. You know? It's it's you know, I'm not gonna read out all of these kind of challenges, with AI, but I guess it's not out of the box. And if you don't have your stuff organized and a plan and policies, it really stuff ensue me all the things that I I suspect you still advise your clients before you they just start slapping tech on all their problems to to kind of plan things out and get yourself sorted. Like, you can't have a great AI experience with a lot of bad data or unstructured data. So is is this the case, or, is that is it really just, am I mistaken? You slap some AI on it and everything will work out just fine. In in minuscule areas, maybe, but, not not at at scale. You know, you're really gonna need to have people process then technology. It's it's what we believe in. It's and I think this is just another piece of technology that needs to be leveraged in that way to get scalable results. Yep. Alright. So so I guess with that, you know, let's maybe talk about how to to not choose the wrong starting point. Right? Because, you know, I think with AI, this is a lot more common now. Everyone's starting with, hey. I've got solution x or a couple of AI vendors that I want to just start using tomorrow. Or, okay, we already have Copilot. Let's just use that for something. And and then try to just start randomly solving problems. But it does turn out that, you know, well, the statistics kinda show that then ROI becomes elusive. Right? So Mhmm. You know, is this is this something that you've seen play out? Like, I mean, I know there are great success stories as well. But if, you know, with that unplanned, let's just buy the tech and figure it out later. It's probably not the right move even if it seems easy enough to to deploy. Right? Yeah. You can be like a broken clock and maybe get it right occasionally, but, the reality is you still need to identify the pain points. And so what we even have a a formal exercise that we run with clients. Remember the old fashioned four box exercise, the urgent important matrix. It's called the Eisenhower matrix. Right? But we we typically run an exercise like that with with people and tell them, just track don't track time. I know attorneys say that, but track your tasks and just, you know, what are the things you're working on? Stick them in one of the categories. If you wanna even we've even done our our kind of a rapid, response version of that where we'll say for one week, just track the 10 things you hate doing in your job. What are the things that are driving you crazy? Are you, you know, chasing signatures or stamps? Are you, you know, trying to take information from one form and fill out another form? Is it expense reports? What is it that is is kind of driving you nuts? And then you, you if you simply, you know, track those things, then you can find an actual AI solution that will solve for those things usually. And so what we'll see is, like, sometimes it's a a chatbot that just answers the same question over and over again. Sometimes it's an agent that will will actually, you know, fill out forms for you or what have you, but you start with a problem. Mhmm. Love that. You know, what what I'll say before going to you, Sumi, is I know, when I think about this particular issue, I've heard a lot of, you know, kind of issues with pilots or and I think AI experimentation, not implementation, they're one and the same to me in that a lot of organizations and I wonder, if any members of the audience are doing this as well. And please do ask us questions too, in chat. You know, Steph, Sumi, and I will try to answer as many as we can, in the time we have as well. But, you know, people are trying out AI. Let's use it, but in a very experimental haphazard way. So you're not necessarily solving problems, and you're not necessarily able to prove ROI, but it takes up time and resources, especially for organizations where everyone's busy, but we're also kind of playing around with AI to figure out what we can do efficiently, hopefully. You know? So that that's something that I'm seeing a lot in conversations with, you know, other legal ops professionals in the, what, 500 conferences that I get shipped out to this year. So, Sumi, any any thoughts on on, like, I I guess, this and people choosing kind of the wrong starting point when it comes to EFA? Yeah. I mean, I I think it kinda goes back to what I was saying around, like, figuring out a policy, figuring out, like, what your structure is, and then going into, you know, the experimentation. Experimentation. I I think there is a there was a time and a place for experimentation. People needed to understand what it was before even coming up with use cases. So I will say there was some practical benefit. But, you know, Steph knows, we we have clients that have purchased everything under the sun and used it and continue to use it. So I think that's the the thing is, like, well, what is it you know, going back to what we've talked about before, what is it that you're trying to solve? Mhmm. And it's not unheard of to do pilots. Right? But make it a focused pilot to understand, like, does it meet our business requirements? And if if it does, great. If it doesn't, move on right to the next thing. But, I think having some having those that experimentation being unstructured Mhmm. Risks you having pockets of everything, in your in your organization, and that's gonna lead to you know, you're gonna have a real trouble trying to capture any data around that. Mhmm. How you know, one of the things with pilots because I I like the idea of that the control pilot. Let's try it out, but for very specific use cases or pain points. Mhmm. What about the prework that goes into that, Sumi? You know, just, you know, figuring out what to even try. Do do you have any advice for the audience? And, you know, hey. We've got a lot of problems, contracts, too much manual work. How how could they go about kind of figuring out what the fastest, you know, meeting? And I know we're gonna touch on all kinds of use cases in a little bit as well. But, you know, for someone that just goes, I have no idea. I just know everyone's complaining, that legal is slow and we need to be faster. How could they even begin to get started? I I think, like Steph said, just documenting some of that. It what are those use cases? What are the again, going back to the basics of what are your key business requirements? Take out the technology for a second. It could be, a CLM. It could be AI. It could be an e billing system. Right? Like, what is the problem that you're trying to solve? Mhmm. Get a little more granular than a top level. But, like, critically, what would I, you know, dream state, what is it that I really would like to do? What would make I hear that I'm going really slowly. What is it that would make me go a little bit faster? What are the resources that I wish that I had? Starting to think about that will be helpful. And then you can maybe partner with your IT team to, you know, understand capabilities. Is this something that AI can do? I mean, we are, at the end of the day, lawyers, so I don't think we're going to be tasked with trying to figure out, like, the tech true technology and what it it's, full capacity, but partner with people in the organization, to understand what they're doing, what they're using it for, what in the industry, what are peep what are your other colleagues using it for? And it kinda jump starts that use case component. But, you know, at the end of the day, it is really just, like, what is it that you, actually need? I like it. And, Navan, one one point of the the POCs too that I was thinking of as you both were talking is that the if you're running POCs look. We we like Sumi said, we work with clients that literally they're now running POCs, meaning not formal. It's just attorneys using stuff that they they kinda got permission to do. There's, like, seven different things going on in the organization. And if you're doing that, it's really hard to control it. It's really hard to compare apples to apples. Right? Like, okay. Well, what tool is actually working for what we need? And so, you know, I think part of the legal operations mandate really should be to kinda lock that down, really understand what problem is the the problem that you're trying to solve. Is it redlining? Is it finding things? Whatever it is. And there are probably different tools for those different use cases. But what I was gonna say is we have we had a client who really did it well. And what they did was they locked down the POC so that if you wanted to get into this pilot group and people were kind of there were, not everybody, but some people were clamoring to get into the pilot group. As part of it, you had to, calculate the ROI. And so I know we're gonna talk about ROI more, but I think, you know, really being really saying, like, look. This, I I would before I was doing this one thing, it took me three hours a week. Now I'm down to fifteen to twenty minutes. That's the ROI. You don't have to do the math or anything, but to calculate the time savings. And so that's part of being in the pilot group is capturing that ROI. Oh, I like it. I make it exclusive. You know? So Yes. And plus plus, I mean, you're too big of a pilot. It just starts to become noise or, you know, let's call it unstructured data. Right? But when we touched on this already, you know, starting with pain points, starting with use cases as you know. So I I should have advanced the slide, but I got way into the conversation with you and Sumi. But but, you know, it's my key takeaway as always kind of been, you know, use the tech where I can get the biggest benefit. Right? And that comes from, like, you know, figuring out what use case and what problem we're trying to solve. You know? Just just because there is, let's say, one VP of sales because it's always a VP of sales. Right? Making the most noise about a couple of, you know, customer contracts, but that not my that might not be the thing to actually solve because once I start looking at the data, is if that's three contracts out of, you know, a thousand a year, that might be something that is much better suited, like the admin work of legal, you know, monitoring email and stuff like that. And so I do wanna belabor this, but, you know, I I like the theme of, you know and I think, Sumi, you touched on it, kinda using AI to to solve specific problems really, really well instead of just general AI use. And and so I wonder, I guess I'll go to use that since I I called Sumi out already. It's just, you know, when you think about specific use cases, is that the trend that you're seeing? Identify your top three most efficient problems to solve and then go for those, or is there a better way? I think that's a good way to start and, but knowing that it's a little bit of more art than science here. And so sometimes where you wanna start is with your biggest champions. So it might not be the biggest bang for your buck, so to speak. Like, it might be a use case that only helps five people in the department versus the whole department. But if you've got people who are really revving to go and are very excited about it, then that could be a good place to start because you you're gonna get you have instant momentum. And but but I think the the story that you wanna be able to tell is if you're doing a a small pilot like that, what you don't wanna you don't want it to seem like, oh, it's just gonna help these people. You have to still tie it to everything else that's going happening across the department. Like, we're gonna pilot it here. It's for the specific sorry. This specific use case, but, that's analogous to, you know, all these other things going on in the department. So if it works here, we're also going to expand to these other departments. So I would say yes. Of course. Look for your top three use cases. That's a great place to start, but make sure they're manageable. You don't wanna bite off more than you can chew coming out of the gate. So we always say, you know, crawl before you you walk and run. So make sure that it's small and and manageable and kind of, something that you could implement on, over time or you know what I'm saying? Build on over time. But then also look for the people who are gonna be engaged and active participants because that's gonna be really critical too. Awesome. We we've actually got a question, from Joseph Cascarelli. Joseph, sorry if I said your last name wrong, but, will AI negatively impact on attorney hourly billing? You know? So doing research that used to take one and a half hours, maybe taking fifteen minutes now. I want to I'm a self confessed billable hour hater. So, you know, my guess my feeling is, you know, from a law firm perspective, I know a lot of the big firms are embracing AI and trying to deploy it and use it as well. But maybe that's moving to fixed rate pricing or, yes, you know, there will be a reduction in hours, and that just means being able to support more clients. I know, Steph Sumi, you both have lots of thoughts around billable hours, as well. But any any response to Joseph? Look. I really hate the billable hour, Navin, as you know, and I've been calling for its death for twenty years, and I've been wrong up to this point. But I think that, yeah, I think AI is gonna have to make law firms think about rethink their model. I know it's really complicated because I know compensation is built on that on that model. So it's it's really gonna be a whole reinvention for a lot of law firms, Not all of them. You're still gonna have, you know, the top tier law firms who are gonna be able to charge whatever they want, and and legal departments are gonna pay that. But for the majority of law firms, I think you really do start to, you know, need to start thinking about moving towards, alternative fee arrangements. Mhmm. We work with our clients now to move off of hourly onto flat fee arrangements. Mhmm. It's and and by the way, we don't we don't want the lock rooms to lose money. That's not what this is about. It's about predictability on the in house side. So when you're in house and I've had to sit my butt in the CFO's office and explain why we were over a budget in a quarter. And let me tell you, that's not a fun conversation to have. But, you you know and what is there to say besides we just had no idea the bill was gonna be this big? And that is not a good story. Your head is on the chopping block when you have to have that story, that conversation with the CFO. And so for us, it's about building and predictability. If you have monthly retainers, you know, some months we win, some months the law firms win. But it's it's a fair you know, at the end at the end of the day, it's it's you hope that it's a fair engagement for both parties. So, really, it it I'm hoping that we could move more towards those types of models. We have a pricing expert on our team, Navin. You know, Ken Callender Oh, yeah. Who does yeah. You know Ken well. He, he negotiates with law firms and moves to value based pricing, which is really, you know, for every phase of a big matter, this is what this is what the law firm is gonna charge. And so I think law firms have the data. They can do this. And and I think sometimes, you know, what's interesting is they're I know I've spoken to a lot of law firms who are like, we tried, but then they want us to shadow bill and so that they can see that they're getting a deal or they're afraid to try, you know, because they think that we're gonna rip them off. So we culturally, I think we still have a long way to go on this. But, yeah, Joe, I don't think there's any way to look at what's happening and to to think that this is not gonna have a massive effect on on the hourly bill. Like you said, when something used to take you five hours and it's taking you a half hour now, that's gonna affect the billing. Yeah. And I I think there's an opportunity to get creative here in terms of, you know, the push from your CEO around show me ROI for using AI. It can expand to just beyond your in house team. Partner with the law firms to start thinking about, hey. If you can show me that ROI, it gets me more budget later. And it it's a full story as opposed to just, I just need a, like, a smaller bill tomorrow. And I think they get they will benefit too because it shows this great partnership. You know, everybody's always talking about I don't wanna ruin the relationship. This is a relationship building exercise, and a way to frame it in in a different light, than just, like, cut 10%. Right. Well, I know I know everyone. We've kind of gone a little negative sounding on AI, but now we're gonna talk, in a second about, the right approach, some great use cases that that all of us are seeing. But before that, our next poll so and I wanna bring in, the question you posted in chat as well, Mark. But what's your biggest frustration around AI that you're facing today? You know, is it that I can't I don't get to use it at all? Is it concerns about data privacy? Is it something else? Let us know. And and I know stats soon. Data privacy comes up all the time when it relates to AI. Right? Is my data being, you know, processed offshore? Is it being used to train the model without my permission and stuff like that? I do kind of feel that with most AI providers in the market today, we've come a long way from that fear. I mean, you can negotiate provisions into your contract. A lot of this has become standardized. Is this is this concern kind of maybe overblown or is it still are you seeing it still be an issue, in the market today? I I think, that certain parts of it are maybe we've, you know, understood it and resolved for it. But, you know, the static AI technology is not static. It's ever changing. And so, the onus is on the team to start to understand, like, what is what is actually happening. What is agentic AI? You know, we talked Steph said earlier. You don't want it, like, running furiously around on your server or computer. So, like, understand what it is before you start to, like, say, okay. The these are the issues. You know, and I think there will continue to the issues will continue to evolve as the technology evolves. But, yeah, I think for now, at least the baseline, what you said, Evan, is is right. You can you can certainly, negotiate around training the model and things like that. Alright. Well well, thank you everyone. Well, thank you, Sumi, and, thank you everyone who answered as well. And now we can move into, I guess, the fun part or the happy AI part, we hope, of this session as well. You know, people, process, and technology. Sumi, I know you touched on this, a little bit. And and maybe it relates to, Joseph's follow-up question really around, you know, hey. What about when we use AI to research and the AI gets it wrong and, you know, it impacts your client relationship or a judge calls you out. And I think that this goes right back to people are still really important, not just kind of as stakeholders, but depending on the AI use case, still being that human in the loop. Right? Is that is that, like, kind of am I articulating this in a sensible way? I mean, I I would treat the I would treat AI like a a young a very junior baby lawyer. Right? And so you would you you you're not gonna send anything to the client that you didn't review anyway, and you would be charging for for that review of that work. But it's a lot faster. Right? You're not doing it yourself or you're not paying an associate or, you know, five hours or whatever to do the work. So, yeah, you would bill for that because you would you would be re you would be billing for any review work that you did. Yep. A 100%. I will always think of, you know, AI research tools, right, as, you know, case law on steroids, if you will. Right? It gets me lots of answers a lot faster. I still need to check, you know, if I'm gonna start bringing into it. Right? So at the end of the day, I think as as the attorney on file or if you're in house as well, you know, you still want to double check, and most people do it. But like you were saying, Steph, you know, if it's say I had a junior paralegal that did some of the work, You know, I checked especially if they just joined the organization. Right? Perhaps after after working with them for a year, I kind of, you know, put place a lot more trust. So I think, again, with AI, there's a a trust but verify specifically when it comes to legal work. I'm not really talking about administrative, like, lift type of stuff. Right? So so I know people process technology is is something of a a catchphrase that that Uplevel uses a lot. Right? And so, maybe if I could go to you, Sugi, you know, because this really resonates with me about have and I don't even think it's limited to AI programs. This is just when you think about technology for legal and legal operations. All of the above still applies in many cases as well. What does this really mean, and and what should the audience really take away from this? When we talk about people process and technology, really, we're saying, like, are the right people doing the right level of work for people? Right? And so, I mean, that is a critical question when it comes to AI. Like, are are, you know, expensive attorneys being reviewing NDAs basic NDAs? Are they, you know, filling in templates? Are they, doing, like, manually updating trackers? All of those things are, in my humble opinion, not something that you need to train as an attorney to be able to do. Right? And so it's things like that you start to understand, like, okay. Let me take a look at people. What are we doing? And then what how are we doing it? Yes. There with technology, it may, help you, create efficiencies. But but if you have an inefficient process, it's still just gonna be a inefficient process. But maybe parts of it are faster, but it still creates this pure unnecessary bureaucracy. And so trying to understand, like, all of these components, to both figure out what your use cases are and where you can really leverage the technology, in a in a really valuable way. Love it. I love it. Steph, anything anything you want to add perhaps publicly contradict Sumi? Not on this slide. Let's see if there's another. I think she's right here. No. No. But but I I think, you know, scaling technology efficiently. Right? If it's just me using the tech, it's fairly easy. I could just spin up, bunch of stuff with Claude and have a good time and and, you know, not worry too much. But when you're really trying to deploy successfully across a large organization, it's it's not as simple. And so 100% agree having the right people involved. And and I know that takeaway of just do you even want people doing this work to begin with? Is it essential? Answering all these questions is really important. What I always fixate on is process. Right? You know, it's kind of the we used to say this when it comes to CLM or any workflow release. Yes. You have your process that you built, like, calculus on the back of my teeth, over the course of ten years with all these steps and all these approvals and all of this. And then if you take that and dump it into an AI agent, is it really helping? You know, it's streamlining and making it more efficient. I presume that's something you're seeing a lot of, right, in this space when people think about, let me automate with AI. Don't just automate as is. Like, look at that process and get all the dump out of it as well. Yeah. I think this is You yes. Sorry. Oh, yeah. I I it just one thing comes to mind is that so we are working with a client, and they did want to use AI to start to, you know, accelerate their intake process, and review materials. Because a lot of materials were coming in, and they were not, you know, full sun. They didn't, adhere to policy. And then, you know, the box starts working and it's showing you what what all the policy, flags are. That's the point where where then people started to look at and go, oh, okay. We have a lot of flags here. Are they really necessary? And so you can kind of work in tandem, with with the technology to highlight. Because sometimes it does. It highlights really where your process inefficiencies are. Mhmm. And now the client is taking that back and saying, okay. Let's take a look at, one, we can effectively, triage this policy because now we know, like, what's coming up over and over again as opposed to just rehauling it like, overhauling it, you know, and taking six months to do that. They can pinpoint these are the things that we need to fix and then fix them and then keep moving. So, you know, there there are a lot of really good ways to to use it that way. No. Thank you, Suni. And so, you know, we can we can advance what I can start addressing Steven Zersky's question about, you know, reviewing the AI review because what we're gonna talk about is some AI use cases around contracts and contract intelligence. I know that one really common use case for AI is reviewing contracts or redlining and drafting and negotiating contracts in something that skilled attorneys have spent years building up a a career around. And I always think of AI. It's really two use cases. One is AI excels at pattern recognition. It can find stuff in a document. Not necessarily even a contract, but let's just talk about contracts for now and tell me what's in the contract, like highlighting it. My I'm dating myself a little bit, but kind of it's an accelerated cool tech version of those pieces of red tape and stuff that people would stick to our paper contracts to highlight certain sections that, you know, we need to look at. And so I think AI does that really well, like, identifying specific risk flags, and you can build your playbook around, we don't like this, we don't like that, we, you know, want the contract to not be assignable, Blah. No auto renewal blah blah blah. And then when you, the attorney, pop it open and run, your tool of choice, it will identify all the places where, you know, hey. This is bad. This is good. This is fine. And then now you have that second part of the use case that you don't always have to use, which is redline this for me. Maybe it's AI redline some of these for me, do the whole thing, and then now you can check it. I know personally, I don't like checking other people's red lines. I like writing my own. So when I would red line a contract, I would rather have all the issues identified, and then I'll make my own edits with my style and my flair and whatever. I know, Sumi, you you redline, a contract or two, in your time. Do you do you have any thoughts around, you know, the AI redlining use case for contracts? Yeah. I mean, I think it's getting better and better for sure. And, look, I don't think that you can use out of the box LLM right now to do a lot of these, what we're gonna call advanced use cases. It's an advanced use case. Can it be done? Yes. Can it be done accurately? Yes. Are there tools out there today that can do that? Yes. But, like, you you know, I think to answer, the question earlier, which is, like, do you have to use multiple tools to verify? No. You just have to find the right tool that can that is designed for your use case and that, you know, that's what the pilot is for. You're testing out to understand, is it getting what I need? If it's not doing it, don't use that tool. Right? Similar to, like, just an e billing system. If it's not going to be able to handle your VAT calculations and your global invoices, like, don't use that tool. And so I think that there are tools out there that are very bespoke, for particular use cases, and it can do them pretty well. But you just have to, like, experiment a little bit with and find those. Do I think that everybody should be using all of those? No. Because I don't know that that's necessarily everybody's pain point. If that's your biggest pain point, then invest in in trying to understand that piece. Yeah. And to comment on the on the part of the question that was, you know, said it seems like a lot of work, right, to if you have two AI tools reviewing, and then I have to review it anyway. And I I think that's what Steven was saying. But Mhmm. The like Sumi said, if you get the right tool and implemented it properly, it it should work without any secondary AI tool. And, again, these tools are getting better literally by the day. But then, you know, I guess the scenario where the CEO hands you a contract and says, have at it. You're look. You you should there's gonna be some instances where you're at this you know, like, we used to say, we're not negotiating NDAs anymore, but then one of the lawyer an NDA. They're reviewing it. Like, that's just the nature of the beast. But for think about all the other work coming in. I don't know what your department is structured. Like, maybe you're a solo GC. But if in a department where you've got, you know, 15 or 20 lawyers, that the amount of work that's saved by just looking at the flags, you know, that the that the AI is identifying, it's a massive amount of work that you're actually saving. And it happens in seconds. So to say two AI tools plus human review is a lot of work. The that first review that you're having done by AI is zero work. It's no work whatsoever. It happens in a matter of seconds. And so yeah. Think of it like that. Yeah. Not not to beat this this review dead horse, but I think definitely you're spending time on this as it relates to contracts because that's that I feel is where the market is focused, and everyone has some variation of an AI redlining tool. I always think about, you know, back when I had to negotiate, you know, 50 or 50 page or contracts in the hundreds of pages, being able to find that stuff quickly, is a huge time saver. And so, again, it goes back to what you were saying earlier. So well, I know Steph, you said it too, but, finding the right pain points to focus on with AI, I think it's super powerful. And I'll share something. I was gonna share it at the end, but so Agiloft is launching, a tool next month as well called Astra. It's exactly this. It's contract analysis using AI. So you can build or download playbooks and then just apply it to a contract you're looking at. Let me find my 10 high risk positions or I'm just reviewing a statement of work. Have AI just confirm that none of these particular provisions have been added in. Because, you know, when we review SOWs, it's always, let's make sure no one's trying to amend the master agreement secretly in here. You know? And so looking for those things and just confirming they're not there, that's something that AI is incredibly powerful and accurate in. And then when you ask it to, hey, generate some language for me, that's, I think, where you want to, you know, still put some eyes on it. Before I leave here, I know contract data is a big thing that especially, I think, with procurement teams or anyone that has to really deal with contracts after they've been signed, as well. And, Steph and Sumi, do any of these use cases in particular resonate with you around, you know, the contract repository or legacy agreements when you work with your clients? Yeah. I mean, these are all good ones. You know, I think these the the renewal exposure, you know, one of the things that we've always said is, you know, for especially for legacy and payments, you know, you're gonna if there's an issue with the contract, you're gonna wanna read it. That's the the be all end all. If there's something, like, material the if the situation is material enough, you're gonna look at it. So, you know, tagging all these things is is, like, good for general data analytics, and I think it can help you position yourself and gives you more of an insight into your repository as a whole. But, you know, at the end of the day, are you gonna rely on somebody tagging it whether it's AI or a human? Probably not. You're gonna wanna you want, like you said, to be, you know, kind of taken directly to the to that provision so you can take a look at it. And then that's that's time savings. And now you're the part of your time that is being spent is really the strategic one. It is using your legal expertise to take a look at, that contract, that language, and develop an argument for or against whatever you're willing to do. Alright. Perfect. Thank you. And I'm gonna breeze through some of the next slides because I really want to spend some time on the two case studies that, you know, you and Steph are gonna share, as well. But we're not gonna not talk about legal operations as well because I know Lauren, also asked the question around, you know, kind of using AI to connect different domains, you know, procurement, commercial, m and a, IT, and stuff like that. And and I think I look at it almost in the context of using AI to create this massive data layer for legal operations as a whole where, you know, you can then draw on, you know, information that's tailored to different departments. Let's call it making dashboards. Right? Because we all know corporate, folks especially love getting lots and lots of dashboards because it's easy to digest that information. Is that something and and, if it's something that's in the case study that you can talk about, we can skip ahead to that as well. But from a legal ops perspective, stitching these different domains together using AI, Is that is that a good use case? Yeah. It's go ahead. Yeah. I think, you know, it it's dashboarding. I'm chuckling because it's been the bane of my existence, and Avenue and I have worked on dashboards together in our in Top Lives. But it's, it's so difficult to get, you know, data from because, Lauren, the question you're asking, the problem is that they all have different use cases. Right? So, theoretically, completely different tools that they're using even though it's AI driven. I mean and so when you're so in theory, yes. You should be able to have AI sit over whatever your system of record is or a a, you know, some, you know, repository and be able to get you information and build dashboards. And I do think we will get there for sure. I just haven't quite seen it in action yet, and we have a team who's playing with exactly this stuff right now to try to see if we can, you know, build out exactly what you're talking about, Navin dashboards that are feeding you're getting data fed in from all of these different areas. So I would say we're probably on the cusp of that versus the reality of it. Yeah. There's I I do have a client that is working towards that, you you know, getting their Salesforce data, getting their CLM data, into their data warehouse. I think that is gonna be still the the objective is to get it into that data warehouse, and then you can apply, AI on top of it. But Steph is right. Like, in the disparate systems, it's still, you know, just you're still having to aggregate the data and, you know, put it somewhere. So I think effort should be spent certainly to create the data warehouse and then, yeah, absolutely, use, AI to slice and dice it. Yeah. So back to our point about readiness work. I mean, that doesn't go away. It's still really critical. No. Exactly. I I kind of I know I heard variations of this quote. Right? AI doesn't always know what's true or what's accurate when it comes to data. It just works with the data that it has, especially if you restrict it from using Internet data, which you really should. So so yes. And I for those of you that answered the poll around contract data, thank you for that as well. And, why don't we, Steph and Sumi, jump straight into your first case study, around with that nonprofit organization that I know you shared with me, but now I hope one of you can share with the audience because this was really interesting. Yeah. This is a nonprofit organization who, Ken, on our team who I was mentioning earlier, he sits on their board. And as you all can imagine, you know, they are strapped for cash. Right? They this is not, you know, they they wanna use every dollar that they get in donations to actually, help the the they it's it's a, an organization that helps, young people, youth. And so they have, we met with them. Brandy worked with them really closely. They have a ton of administrative use cases, not necessarily legal specific. Some are legal focused, but a lot of them were very administrative, like anonymizing files because they're they're working with young people, so they have to anonymize data. There can't be identifying, you know, information between certain files. Right? Those kinds of things. Plus, you know, they were they were actually going through grant applications. You know, basically, think of it like an RFP. Right? If you're completing an RFP and you're doing something over and over and over again, taking data from one information and, or taking data from, everything that you've done in the past and now using that to fill out a brand new application, RFP, whatever it is. So there's a lot of administrative things that this organization is doing, very similar to a lot of what legal organizations or law firms are doing. And so, Brandy worked really closely with them to deploy a bunch of custom bots. So that's basically what she was was initially working with them on. And they were able to save, you know, an additional in administrative fees, $273,000. That money is now literally going to the kids that they that, you know, they support. So that was a massive win for them. Their board was just absolutely thrilled. They are now moving into, more of an agentic, approach. So now they're ready. They're pretty savvy. This this whole team was willing to you know, Brandy was creating the bots for them initially, but then showed them how to how to do it themselves. So a lot of them started taking this on themselves and and building, you know, custom tools in house, which is, by the way, I think, the model that we're all eventually gonna be working towards building our own custom tools in in these LLMs. And now they are ready for agents, so she's on another project with them to help, build out some agents that will actually help them to get this work done. That's that's awesome, Steph. I what I really love is which doesn't really have anything to do with AI, but just the savings and how people can focus on the mission, you know, for something like this. Because, I mean, candidly, it's it's hard to get excited if I save a bunch of time, and I'm now using the time to just go to more meetings that could have been emails. Right? So it's it's really meaningful, but the thing that I think I'm taking away from this is that focus on the operational work. You know, it's not I know we've talked a lot about using AI for legal thought, really, just what, hey, review and redline my agreements is. But there's so much operational work that is just literally mechanical that an AI would be perfect to just absorb so that we can do fun stuff as humans. You know? Exactly. I mean, let's be clear. None of this work was they were not excited to take information from one form and fill out another form. And Sumi and I, that doesn't get out of get us out of bed in the morning, like, yay. We're gonna go save a corporation money. But when we could actually say when, you know, when you that's why I said initially, like, start with what you hate doing. Because if we could look at, you know, somebody on a legal team or or a law firm or whatever and and, you know, take away something that they can't stand doing that's really they feel like Sisyphus pushing that rock up the hill, and then we can solve for that, that actually really does make us happy. And in this case, this was mission driven, and just those administrative costs that they cut now go to the kids that they're helping, it it did feel really good. Start with what you hate doing. I think we could end the webinar just off that line, really, as as a catchall. This is the way to get ROI, you know, very quickly with AI. Yeah. Let's move to the next case study. I know this has a really big sexy, you know, number, attached to it as well. And and for the audience, you know, we we will provide a QR code at the end, and you'll get the slides too. But there's a bit more detail around these case study that these case studies that staff and Sumi have very kindly shared with us. But I guess one of you tell us a bit about this global insurance organization that that you helped as well. Yeah. Well, I think it's it's a funny story. Right? Because we we have been working with this team for a very, very long time. And they started out, you know, fairly immature in their legal operations and have have slowly taken steps. And this is this is one of the things that we talk about a lot is, like, the phased approach. And they really kinda took this and, piloted this. This is the one step we're talking about earlier where they, asked people who wanted to be in the pilot, and they were then have had to document, the the time savings that they had. And I I think, Steph, you had mentioned to me a while ago about the, the GVC got so used to these people working so quickly, the people in the pilot, that when she had to wait for something, from a from a person who wasn't in the pilot, she got frustrated. She was like, why is it taking so long? And then had to be reminded, oh, yeah. They don't have access to these resources. It was so it was a really stark contrast in the end, and it wasn't had nothing to do with the the, individual's ability. It was just, like, the efficiency was driven way up, because of access to these resources. And I know we talk about hours saved and, you know, FTEs saved. But I think going yes. That's all great. But it it just means, like, you can maybe keep the resources steady, but get them to do more interesting work, and, you know, be able to make decisions faster. You have more you can you can synthesize data in a way, that you couldn't before. And I think that's kind of the interesting part of it, You know? Because a lot of times we used to talk about, I wish I had time to look at this and and analyze it. I would be able to make a, you know, decision on policy or on structure or on a deal. And now you have access to those things, that, yes, maybe a data scientist in your organization could have pulled together for you and it takes them many weeks. You could do that in a matter of seconds, which is really kinda cool. No. Definitely. Yeah. Oh, sorry, Steph. Go ahead. No. I was just gonna say to give a little more, information on this one. It was 16 people. They're a team of about 80 people. It was 16 people in this pilot, and the the general counsel, also IT in the middle is kind of fun a funny story. Tried to cut off access to they're they were using chat g b t. IT decided they wanted to move all towards Copilot and cut off access to chat g b t and which meant the bots weren't gonna work. And so, the GC got so furious. She put her she made herself head of the AI committee to push it through. And she's like, absolutely not. We're not shutting this down. So she went from, like, being, like, meh, okay, to, like, such a believer. She was such an advocate. And she equated this to, like, being able to hire four extra attorneys on the team that she was not gonna be able to hire. They weren't they absolutely weren't gonna get those heads. And so and so, like, Udemy, I think I'm saying your name. I I hope I am anyway. And the comments is saying, yes. That's exactly the point. This is we are not automating fun stuff here. We wanna automate the stuff that's grinding you down so that you have time to work on the fun stuff. No. Exactly. I I mean, like, maybe I sound arrogant when I say this, but I did not go to law school to check NDAs or, like, check a 100 NDAs a month. And so it it's it's one of those things where I know NDAs, like, the the topic we always beat up on when it comes to, you know, inefficiency and legal, right, stuff it's to be. But it's that could be a perfect use case to just hit low risk, high volume legal documentation. Have the AI check it and then look for his slides if it's fine. In fact, one of our customers, if any of you are in Dallas at the end of the month, we'll actually be co speaking at legal ops dot com regional on the twenty fourth. So come join us, Susan Zikorski. But they use AI to just triage contracts. You know, so all the low high volume, low risk contracts. Look for these, you know, specific risk flags that we've agreed with. And just if it's not there, send it to everyone else, procurement, finance. They can review and sign up. We don't need to waste legal time. And that lets legal look at all the cool fun stuff that we actually want to negotiate. So we're at the end, everyone. I've got, just two or three quick things to share. But Steph, Sumi, thank you very much for joining us. I guess, Steph, let me go to you. If you have one takeaway for the audience, if you're thinking about, you know, using AI and are a little bit anxious or don't quite know where to start, aside from call up, level ops, and, you know, we have continue that discussion, is there is there anything you want to kind of tell them to go to? Let's say on Monday, you know, it's it's the Yeah. I mean, like, if, you know, you said you wanna kill that phrase, AI is gonna replace lawyers. I also wanna kill a phrase. It's the do more with less. I think there's nothing more demotivating than that. And so I think just playing with AI, you can do something really simple like create your own styled writing bot. It's really easy to do that in chat g b t. Train it on your own writing style and just start playing with that. That my writing bot has saved me hours. Oh, it saves me hours per week, honestly, because it writes just like me. So start really simple, start there, and just start playing with it. And, yeah, I mean, there are people in our networks who are willing to help you. Alright. And, Sumi, last last word. Any any quick word of wisdom you wanna share with the audience? I think, you know, don't be afraid, and you can be a leader in this space. This is one of those places where, in organizations, people are looking for some leadership, and legal is really poised to to be that leader. So, you know, take it and start to understand it and and take that opportunity. Thank you, Sumi. Well, we have our last poll as well. If any of you would, you know, like to continue the discussion with Agiloft, with myself, please reach out. Just, select whatever applies, if it does. But I'd also encourage you to, you know, connect with all three of us on LinkedIn. Steph and Sumi, I I know this for a fact, love talking about this stuff, and I'm sure it came through in our session today as well. So if any of you have follow-up questions, if you have a unique AI issue that you're trying to solve, they are both amazing people to to chat with as well. So, you know, reach out on LinkedIn and go from there. And finally, for all of you that are still here and haven't jumped off to go grab a quick lunch, up on the right, up level ops has kindly, you know, provided us with a QR code. We also have a PDF, that we'll send out. But if you wanna scan the QR code to look at the AI ROI calculator and, information about the case studies you heard about today, please go ahead. And I talked about Astra here in the session as well. If you want to join our wait list, I have a special wait list that the rest of AgileSoft does not. So if you want to scan that QR code and sign up, it it's a contract intelligence and redlining tool as well, and I sometimes dabble in building playbooks for it myself as well. So please be my guest and sign up, and give it a try, and let me know what you think if you wind up doing so. And with that, thank you again, Steph, Sue me for joining us. Thank you, to the audience. I know, Ronald, you had a question around signing up for CLE credits. Jen will, help that as well. So, but that's it, everyone. Thank you for joining us, and have a great rest of your week and, summer weekend. Bye.