LexisNexis CTO Greg Dickason On Building Legal AI In Real Time with Customers in Its New Innovation Lab

0:02Welcome to Law Next. This is Bob Ambroji. This week, LexisNexis announced the opening of its Customer Innovation Lab in New York City, a physical space where legal professionals, LexisNexis engineers, and AI companies such as OpenAI and Amazon Web Services sit in the same room and build legal AI together in real time. The company is calling it a new model for how legal AI gets built and shipped. Instead of developing products internally and asking customers to validate them afterwards, customers bring their pain points into the lab, engineers code against them on the spot, and prototypes that pass muster can move into Lexus Plus with Protege in a matter of weeks.

0:46My guest today is the person leading that effort, Greg Dickerson, Chief Technology Officer of Lexus Nexus Legal and Professional. Dickerson joined Lexus Nexus in 2019 and previously ran the company's Asia-Pacific business, where he launched Lexus Plus AI in that region. Before LexisNexis, he was chief technology officer at CoreLogic. In our conversation, recorded just ahead of the launch, Dickerson explains what actually happens inside the lab, including watching a solution get revised four times in a single hour based on live customer feedback. He also talks about what's already come out of it, why legal drafting is the lab's first priority, and where he sees LexisNexis heading on building its own AI models and on an increasingly agent-driven legal world.

1:39I'm Bob Ambroji, and this is Law Next, the podcast that features the innovators and entrepreneurs who are driving what's next in law. Before we get to any of that, please take a moment to hear from the sponsors whose generosity supports this podcast.

2:17And the next time a matching request shows up on any case, BriefPoint reapplies your objections and response strategy automatically. No data entry, no setup. Watch the 60-second demo at BriefPoint.ai. Make your life easier with Practice Panther, the leading legal case management software. With easy-to-use features such as automated billing, customizable workflows, and seamless client communications, you can accomplish more in less time. Trusted by tens of thousands of lawyers across 170 countries, Practice Panther is a partner you can rely on. Visit PracticePanther.com today to see how you can take your practice to the next level.

3:07Greg Dickason: Now on to today's conversation. Greg Nixon, welcome to Law Next. Thanks, Bob. Good to be here. So as we are talking, LexisNexis is just launching this customer innovation lab in New York City, which the company is describing as a new model for how legal AI gets built and shipped. So let's start there. Can you kind of break that down for us? What's new about this? What's different? What are you actually doing with this new innovation? For me, the biggest thing is we're collapsing the distance between the people who know what the problem is and the people who can build the solution, which is our engineers and our data scientists.

3:50Greg Dickason: And putting them in the same room means that our technologists really understand the problems, which means that our ability to respond very quickly to where the market's going is hyper-accelerated. That's what I'm really excited about. But not only that, we're also bringing in the AI labs, some of the hyperscalers. So we've got three different people in the room. We've got the customers, our clients, the law firms in our legal departments. We've got our own engineers who understand what they can do and what's possible. And we've got the AI labs and the hyperscalers who can really tell us where the future is going.

4:22Greg Dickason: And the combination of those three, I think, means that our rate of innovation is going to accelerate dramatically. So obviously, LexisNexis has a long history of developing legal technology, of innovating legal technology. Why kind of take this different approach at this point? So I'd say if you look at what we've been doing the last couple of years, we've been probably three, four, five times faster than we were at producing products. If you go back five years, we're probably five times faster than we were five years ago. Just in the way we've structured ourselves, in the way we're responding to the market, in the way we are building internally with the agents, and the way we're working closer with customers.

4:58Greg Dickason: But I think we need to go another five times faster, Bob. We really need to get to the point where we understand all the tiny nuances of what makes the biggest difference for an AI solution for our customers or for any solution for our customers, including AI. And that's why we're doing this, because it helps us to move faster than the startups even. So I've sat through any number of LexisNexis briefings over the course of my career. and I think I've commonly heard LexisNexis executives tell me that customer involvement in product development is a core tenant of how they do build and develop products.

5:37Greg Dickason: And so is this something different in the way that you're bringing customers into this? I think it is in two different ways. So first off, how were we doing it in the past, and we're still doing it, is doing a lot of customer conversations and really deeply understanding the problems. That still needs to be there. So we've got big teams that do that to spend a lot of times with lots of different customers. But I think sometimes that refining bit where you get the problem and then you have to refine it to get it exactly the way the solution should be, that can take too long. Because you get the problem, then you spend time building something, then you bring it back and show it to the customer.

6:10Greg Dickason: They give you some feedback, then you go back, build something more, go back to the customer. That cycle can take weeks, if not months. Now what we're doing is literally in the session in the lab, the customers are telling us their problem. We've already got the first solution because we've listened and we've already pulled that together. We show that to the customer. They say to us, yeah, but it would work slightly better if this happened or that happened, and then we tweak it live in the lab. And as a result, in an hour or two hours, we can iterate two or three times over the solution and very quickly get to something that really works for them and then test that with a few other customers.

6:44Greg Dickason: So last week, we saw that live. I saw a solution get changed four times in one hour as a result of customer feedback. And for me, that is the difference, is that final tuning, that final being ready to go live with something that really hits the market, completely solves their problems. That's what the lab's doing. Is the approach also different in the sense that rather than perhaps, and I don't know how this is how it works, but rather than having somebody at LexisNexis sitting there thinking, we think we ought to develop this product, you've got the customers coming to you and say, here's the product we really want.

7:20Would you develop it?

7:22Greg Dickason: That's exactly it. That's exactly it. Because when you've got customers in a room, it's remarkable how much extra feedback you could get. And very quickly, you could see, actually, this isn't going to work. That's going to work. So it helps us to refine and understand their problems. And when you've got the solution person in the room, the person who can actually code, we can rapidly say, this is what I think I've heard. Let me show it back to you. And that's been phenomenal. And this isn't just about the features. This is also about how we build models. It's also about how we use content. It's also about the outputs, the drafting, et cetera.

7:53Greg Dickason: It's the whole lot. The lab is a physical place, not a virtual place. I mean, our customers actually coming in there and sitting down with engineers and product people and others. Correct. It's in New York and we will have other locations around the world. But right now it's in New York because that's where it's so easy for all our different customer sets to pop in and give us feedback at their lunch breaks and the evenings. So we've got ongoing sessions. So it's a physical place. We do think there will be a fair number of, we will send people to our customers and actually do labs with them. And we will have some labs in ILTA as well.

8:29So we're pretty excited about that.

8:32Greg Dickason: That's great. So who gets to participate in this? Great question. Effectively, anybody. If you have a problem you think we can solve, please reach out to your account rep or directly to us. But effectively, anybody. So we've got people who are not even our customers. We've got our current customers. We've got people from CIO level, partner level, all the way down to junior associates at large law firms, small law, solo operators, even a judge. So anybody in the legal industry. And we are going to expand it to other professions as well because, as you know, LexisNexis isn't just about lawyers. So it's effectively anybody who uses our product or could be using our product.

9:14So how do you evaluate these suggestions as they come in and decide which ones to bring the customers in on and work with them on?

9:21Greg Dickason: Yeah, it's such a good question. Yeah. Effectively, we still are using what we've done before, which is all the different insights we have from our sales teams, from our customer care teams, from looking at what's happening in our product. That helps us define where we want to focus. And that, therefore, then helps us to tune based on what people want to bring in. But we're also having blue sky sessions where we bring customers in and we're saying, tell us some parts of your day that frustrate you or tell us the ways in which you think the legal industry can evolve. And in those sessions, we're vibe coding or creating new solutions from scratch and testing those with our customers.

9:59Greg Dickason: It doesn't mean they're going to become immediately product, but it means it's really opening us up to new ideas, which could potentially change the whole way law is practiced or the way our customers get great service from our products. That's pretty cool. I find that probably the most inspiring when you come in and you see people say, hey, what happens if we did this? What happens if while I'm cooking dinner at night, the product can do this or voice interface or something like that? Yeah. and and follow up on that who gets invited question does it matter what kind of firm they're in what kind of legal practice do you have legal departments do you have smaller firms does it go kind of across the range it does and again it depends really on where we focused largely but then we still have open you know effectively open sessions for anybody who has an opinion and wants to come and work with us and that's what's been great there's been a huge huge sort of And almost anybody I tell about this wants to get involved, which I found very exciting.

11:00Greg Dickason: So for us, it's about working out who and when so we can focus on where we seize the biggest areas where we can innovate. Is there any discussion of who owns what ultimately comes out of this or is that an issue at all? It's not an issue. We haven't had that challenge. But effectively, it's for us about how we are progressing our products. So the IP zone by us. And people acknowledge that before they come into the lab. Now, that said, the lab is also set up so that where law firms, particularly some of the larger law firms, they want to explore how they're building their own platforms, how we work with them together.

11:39Greg Dickason: And so the lab's been set up in that way as well. And that's then obviously a different conversation. That's where we are looking at potentially helping law firms build their own internal systems where we can plug in and help them. And that's separate again. Okay. So you're either building with an eye toward enhancing protege or existing LexisNexis products or maybe separately working with a law firm to help, as you said, its own processes develop its own workflows or something like that. Yes, because I'm sure you're very well aware, especially some of the largest law firms are looking at how their approach to AI is quite specific to them, highlighting their skills, their capabilities.

12:24Greg Dickason: And so often they want to build themselves or do part of the process is building themselves. Yeah. I know according to the press release that I saw in just hearing you describe it, obviously the lab has already been an ongoing concern for some period of time already. Can you give me any examples of what's already happened or come out of this lab? Yeah, sure. So I'd say probably the two biggest things have been the vault. So vaults are all where customers can store their own content. And we're looking now at multimodal vaults where we can store anything and at pretty much any scale. And what I'm excited about is the ability to analyze videos, for example, that came out of the lab as an idea and we worked through and it's close to going live.

13:07Greg Dickason: And in that case, you can do something like, I've deposed a witness. It's a two-hour deposition. Tell me when the witness showed surprise and what was being talked about at that time, because that then can potentially be something I can drill into. So our customers can upload very complex content of all different types, documents, audio, video, and then have complex queries across all of those and come up with really interesting insights as to legal arguments they may make or a memo they can draft, et cetera, all based on their content and then, of course, linking back to ours. So that was a piece of work that came out of the lab and has been accelerating off that.

13:48Greg Dickason: Very exciting. And I'm curious, can you at all break down the process for how that happened? I mean, did a firm come to you and say, I do a lot of depositions, but the software that's out there frankly sucks and I wish I had something better? Or do they come saying, I need very specific, I need sentiment analysis. I need to be able to better, I don't know, look for. It was a bit of both. We knew that there was a gap. We'd heard that from our customers. And then as a result of that, we approached a few and said, how do we work? Is this something you want to work on? They were like, yep, let's go for it.

14:23Greg Dickason: And then exactly right. They came in and said, this is the kind of queries we would run. Tell me how you can do that. And then we had a, I can't mention them, but we did have one of our core technology providers in the room as well. And they helped us really refine what was possible with technology. And the combination of those three mean that we can actually come up with a great solution. But absolutely, it was like we knew there was a problem there. But the real understanding of that problem happened when we were actually sitting down and talking it through with our customers. To what extent does this actually happen, like in real time?

14:57I mean, is the customer sitting there and somebody is sitting there with cloud code or doing it themselves or whatever they're doing and kind of trying to build this and come back a little while?

15:09Greg Dickason: The customer looks at it as it looks great. That's exactly it, Rob. Then that's exactly right. So we have a couple of people talking to the customer. We've got an engineer sitting on the side listening and literally writing at the same time, maybe one or two engineers. And then they will turn their laptops around and say to the customer, is this what you meant? and that's what I love is you get the surprise they'll be like wow you've just done something for me like that I've spoken about 10 minutes ago and now there it is and yes do we use ClawCode? That's one of the ways we use, we used different tools, ClawCode being one of them for exactly that reason but it's integrated into our product so we can show it directly inside our product and it's amazing how the technology has moved in the last six months to enable that kind of customer interaction so then what's What's the timing for something?

15:59I mean, I'm assuming that not everything that gets built in the lab makes it into a LexisNexis final product or into Protege.

16:08Greg Dickason: But for those that do, how quickly does that happen? What's that look like? It can be less than a month, but your assumption is 100 % correct. We're testing first. We're not expecting just because you've got one set of feedback from one customer that's going to go live. Often we have to iterate through a few customers. But because it's been built on the platform and we already have working code, it's not that hard to take that from whatever we built to live. Depending, of course, on how deep it goes. If it's content reformulation or a model tuning or something like that, that might take longer.

16:42Greg Dickason: But if it's functionality, easily within a month if we feel that that's what needs to happen. So, yeah. So I'm curious, as you've been doing this, again, I don't know when you actually started bringing customers into the innovation lab. About a month and a half ago was when we first started. So we've had different sessions, and it's ramped up as we've learned what needs to happen. The reason why we're launching it now is because we really wanted to refine what does the lab mean? What's the working cadence? How do we get, as your questions, how do we get what we need to build into the lab? Where do we get those signals from?

17:18Greg Dickason: And then once we've built it in the lab and we've demonstrated it and we've prototyped it, how do we get it back into our engineering teams that they've built it? Because the lab's only 20 or so people, whereas the engineering teams are 4 ,000 or so. So how do you make sure that they're ready to receive rapid iterations? It sounds, from what you said, like you've actually spent time there yourself. Yeah. What surprised you about it? I mean, what's come out of it perhaps or happened there that you didn't expect when you were first conceiving of this idea? A few things. I think probably for me the most has been just the number of people in our business who've just risen to it.

17:58Greg Dickason: What I love is that you get engineers that are quiet, you know, high achievers, but they're quiet. And then you put them in front of customers and they just light up and they start to do things that you just think, wow, that's amazing. And then as part of that is that dynamic when the customer says, hey, I would like to do this. And you see them respond and do it. And then the customer respond. And you get this whole almost creative space in which very quickly you're getting great solutions. And often they're solutions you didn't think would work or you didn't even think were possible. But because the customers describe their problem and because they're sitting in the room and understanding that problem, they're able to build solutions you didn't think were possible.

18:37Greg Dickason: or they've found ways around the technology stack, which makes it much easier to produce what the customer needs. I think often in tech, you sort of get stuck in like, you think you understand the problem, so you think you understand the solution, and often it's quite complicated. And then when you refine, you can say, actually, I can solve the same problem much more easily by doing this slight, simple change. And for me, that's been great, because it also helps us understand how quickly we can build. It's cool.

19:17Lexus Nexus says the lab's first big priority is legal drafting, a market that's already packed with tools promising to do exactly that. When we return, I asked Greg why drafting and what he thinks everyone else is still getting wrong. We'll also get into the build versus buy question on AI models, Whether LexisNexis will follow Thomson Reuters in developing its own AI model, and where he sees agents taking legal work next. But first, please take a moment to learn about the sponsors who so generously support LawNext.

19:58Law firms lose thousands of billable hours each year because the work they do never gets captured. Ajax uses AI to turn activity across screens, email calendars, and documents into ready-to-bill entries tailored to your firm's billing rules and outside counsel guidelines. Entries sync directly to your billing system, helping firms capture over 12 % more billable time. 97 % of firms that pilot Ajax choose to keep it. Learn more at joinajax.com. Don't let repetitive admin tasks slow you down. With Practice Panther, the leading all-in-one legal practice management software, you'll save over eight hours per week with automated workflows, streamlined billing, and secure client portals.

20:52Join the tens of thousands of lawyers using Practice Panther today and get back to doing what you do best, helping your clients. Visit practicepanther.com to discover why Practice Panther is the industry-leading case management software in the market today. Welcome back to Law and Act. I'm Bob Ambroji, and my guest today is Greg Dickerson, Chief Technology Officer of LexisNexis Legal and Professional. Before the break, we talked about the Customer Innovation Lab the company just launched in New York, how it works, and what's already come out of it. Now I want to turn on what the lab is making its first priority, legal drafting.

21:36Why and what does that mean? We'll talk about all that in the remainder of the conversation. Let's pick up there.

21:48Again, based on the press materials I was given before our interview, it talked about that one of the sort of unmet needs, I guess, that customers identified was in the area of legal drafting, document drafting and suggesting that that was an area where they're looking for better products. That surprised me a little bit only because there are so many products in that area devoted to that area. I wonder if you could elaborate on that at all in terms of what you're hearing from customers about that or what you're seeing.

22:18Greg Dickason: Yeah, completely. But I think a lot of it comes down to customers feel like they're getting to 80, maybe 90 percent regardless of vendor, but they're still having to a fair amount of work post that to get it to that 100 % where they're ready to send it out. I'm not saying we'll ever get to 100 % because I think your lawyer always wants to sit down and make absolutely sure that it's what they want. It's in their style and everything else. But the gaps, I would say, is firm style and playbooks, making sure that it is actually written fully to their style, that it's pulled out all the risks and do things like, in our case, Shepard's verify your citations to reduce hallucination and all that, that that is properly highlighted.

22:54Greg Dickason: Another thing we're looking at is context across documents. So I can do one drafting, but have I made sure that that draft and the draft before and the draft before that all lines up? And therefore, there's context that when I read it, that there's context across multiple drafts. And I think that's a gap. It's one thing to do one document right. It's another thing to do multiple documents in order right. Then obviously, there's things like track changes, version history. Big thing for us is trust. How do we surface trust and make sure that they can verify very quickly within the draft that this is good law, that it is good citations, and they can click through and see that on our platform?

23:31Greg Dickason: So there's a bit there. And the other thing I'd add is also, to some extent, analytics around the draft. So I'm drafting something, and I might say this might take you, say, two months to get to trial. And we can run analytics from Lex Machina and other intelligence to say, actually, it might take longer or it might take less. So looking at the draft from the context of the lawyer's workflow and some of the facts they may be stating and making sure that those are accurate as well. So there are gaps. I would say that most tools out there are getting us to about 90%, but I think lawyers want a bit more because it still takes quite a few hours after you've got the output.

24:04Greg Dickason: And that's the gap. Does that make sense, Mark? That makes sense. Yeah, yeah. And is it interesting that LexisNexis has started down the path of building its own legal AI models separate from relying on the general AI models that are out there? Do I have that right? Yes. We've been evaluating models since we did them with Amazon on Bedrock over three years ago. So it's not something we've just started. It's something we've always had in the background. What we do think is that in most cases, foundation models, the frontier models with the right context and the right prompting and something like our knowledge graph and a good rag system pretty much give you state of the art.

24:51Greg Dickason: But there are reasons to pursue your own model. And so therefore we continually do. And in some cases we do use fine-tuned models for certain parts of the legal workflow. Yeah. Do you see that? I know that Thomson Reuters just announced recently that their own model that they have built, that that's actually going to begin to replace some of the general models in the future for certain tasks. Do you see that happening within LexisNexis at any point? I mean, never say never. I would say right now, looking at the absolute frontier models, you know, Opus 5 and others, we still see really good value and quality from that.

25:33Greg Dickason: What we do also understand is that some of the reasons why you're looking at your own models can be from a cost perspective. And that does make sense if you can get the same quality. But we do also focus on making sure that we're not going to pass those costs on to our customers. So we're going to make sure the highest quality is for us the number one bar. And then below that, maybe at some point we will investigate our own models. We already do constantly, but at this point it looks very much like the frontier models. Well managed, in the right RAG framework, with the context graph, etc. I'll give you a cutting edge.

26:08You recently published an article along with the General Council of Anthropic talking about agentic AI in which you wrote, Legal AI is entering a new phase. The first wave proved that generative AI can draft, summarize, and brainstorm at speed. The next wave is authoritative, agentic legal AI that is powerfully integrated, deeply embedded in legal workflows, repeatable at scale, and capable of producing finished, citable work product that meets the highest standard. So I guess both within the context of the lab that you're launching, but also in the broader context of the company, I wonder if you could talk a little bit more about what you see as the significance of agentic AI in terms of legal workflows.

26:54Greg Dickason: Yeah, completely. And I think legal workflows are multifaceted. They go for potentially a long time, multiple months even if you're doing an M &A or a complicated transaction. and the ability for that to be an agentic loop, I think is an absolute breakthrough compared to where we were a year ago. And that's because you can keep feeding back into the model the context that you've just received from the external environment, whether the model's gone off and got deep research from us, gone to the web, got customer content, understood a new aspect of whatever workflow they're doing, maybe new emails come in.

27:29Greg Dickason: All of that context and the loop that the model's doing, for me is what the harness is about. It's the harness is giving the model, this is what's going on now, what should I do next? The model's taking that, calculating, potentially reasoning over that, and then emitting, this is what you should do next, this is what I need to know more of before I can give a definitive answer back to my client, to my customer. And so that loop is something that we didn't used to have before, and now I think that is a super powerful way in which you can use model reasoning with the harness to really get, one, long-running tasks, increase the intelligence of what is there because the model keeps going back and getting more content.

28:06Greg Dickason: And then finally, producing great outputs, Word documents, PowerPoints, Excel spreadsheets, things that the lawyer can use that are nicely all packaged up together and well formatted. And that really comes about because I think harnesses of this power of the harnesses has exponentially increased over the last six months. Are the customers who are coming to the lab or do you expect the customers are going to be coming to the lab, have an understanding of agentic AI? And is that a technology they're kind of thinking about as they're coming in to talk about the projects or the ideas that they have?

28:40Greg Dickason: Some of them are, definitely. I think some of the more sophisticated ones are and they're using Claude Code and they're using Claude Cowork and, you know, the other Microsoft Copilot Cowork now. So they do understand the idea of the agentic kindness because they've seen it themselves. And, you know, we have that inside Lexus Plus with Potager. We have the work product inside that. You can see the agent going through its loop and doing its thinking and demonstrating that to you. So I think increasingly more sophisticated customers absolutely get that. And so they're asking harder questions. But I'd say probably 60 % or 70 % aren't there yet.

29:11Greg Dickason: They're still interacting as if it's a chat. And so there's a lot more to learn for them. And I do see the lab as a great place from that perspective, from an education perspective. So we've got models that we're training constantly in the lab just to show what a training one looks like. And part of that is showing off the agentic harness as well. just so people get an idea of what a harness means relative to a model. But I think, Bob, maybe you're highlighting something that's a challenge for all of us, which is the adoption curve, the maturity curve from lawyers is very wide. There's some highly technical lawyers who completely get it all and are pushing a boundary, but there's a large number who are, you know, it's a background thing and they're not completely engaged in where AI is taking the legal industry.

29:57But I would think Innovation Lab, such as what you've launched here, is somewhat self-selecting in that it's going to attract the legal professionals, whether they're lawyers or whatever else they might be within a legal organization, who are technologically ahead of the curve a little bit. Is that fair to say?

30:15Greg Dickason: That is fair to say. Some of our recruitment, though, has gone and found people who are, we want to talk about drafting. Can you come and talk to us about drafting? And then you get people who are not as technical and you get a different insight then. Is everything happening in the innovation lab related to AI or are there people coming in, I don't know, with project ideas that really have nothing to do with AI at all? I see that happening. It hasn't happened yet. Because you can almost AI everything, right? Any problem, you can find an AI solution. So yeah, at this point, it's been all AI, but no doubt at some point.

30:51Greg Dickason: Actually, no, that's not true. We did have one the other day who was talking about time recording and trying to work out how my product could do time recording. And that's, at this point, got nothing to do with AI. Yeah, well, it could. That could too. Yeah. You know, you also in that article I alluded to, you talk about the fact that LexisNexis has integrated the anthropic legal plugins into Protégé, into Lexis Plus with Protégé. What do you see as the kind of future workflow relationship between like Claude and Protege? I was talking to somebody the other day who said he's got a much smaller legal research, a rather legal product where he was saying that stuff like 80 % of his users are now coming in via Claude, via an MCP connector, not directly through their own web interface.

31:45How are you seeing this evolve?

31:47Greg Dickason: Yeah, I mean, I do think increasingly we're going to see agents sort of disintermediating the user and agents acting on behalf of the user. And we're working on how that works in our system. We don't want agents to directly browse inside Exus Plus with Podge because we already have agents that are giving you really good answers. And if Claude is just getting raw content and trying to pull that together, that generally doesn't work. It's not well-grounded. It hasn't got all of the tuning we've put together. So it generally doesn't work. So what we were looking at more is the MCP or agent-to-agent interface so that that will work eventually.

32:22Greg Dickason: But yes, I do see a point in the future where increasingly people are using their own personal agents to do work. And we have to be part of that ecosystem and play in that ecosystem and make sure that it's an agent-to-agent and we've got the protocol working. It's going to be a fascinating new world, that space. We've been talking about some of the ideas that the customers may be bringing into the lab and developing those. But other than sort of the cleverness of the product idea, are there kinds of other thresholds that you're looking at in terms of developments coming out of the lab that would allow them or prevent them from becoming a product eventually that's incorporated into Protégé?

33:09Greg Dickason: yeah i think there's multiple one is is you know is the interface always going to be you know web web interface or we're going to change the way in which customers interact with us and we're really obviously surface inside like microsoft copilot copilot you know in microsoft teams and other places some of it's about where do we surface and how do we how we there where the lawyer needs us to be wherever they are i think that's one of the big ones the other one I think is that a lot of law firms are reinventing how they see work happening. Who does the work? When is the work done? Does it happen on intake?

33:45Greg Dickason: And so integrating more and more into that, into the processes of the lawyer. So it's not just about what do we change about our product now, it's also about let's understand the day of a lawyer, and obviously that's very different depending on where you are, what practice area you work for and what your level is, but making it much more efficient for you to use our products without you having to come to our products, our products can come to you in different forms. So there's a lot of that going on, I think. And that's why it's great having lawyers in the room with us because you can really learn from them.

34:20Greg Dickason: Not just, hey, can you change this button? But this is what I do in my day. This is what really frustrates me. And then potentially we can help with those as well. You talked at the outset of the interview about the accelerating pace of development within LexisNexis and across technology more broadly. And as you roll out this innovation lab and you talk about rolling out more of them over time, I'm assuming internationally, I don't know if you said that or not, but as you create more of these and you have more and more product customers bringing their ideas to you, how does that play all out in terms of LexisNexis' overall product strategy?

35:00I mean, do you continue to just sort of embrace this rapid acceleration and new product development, or does it all have to reach some plateau or level out in some way at some point?

35:12Greg Dickason: I think it's going to do both in a weird way. What I mean by that is I think the interface has got to get simpler. It's got to be easy for anybody, whoever you are, to come in and use the interface without really thinking about what has Alexis launched in the last month. and therefore it's all about the agent and the intelligence understanding who you are what you want to do and making that available to you but the innovation underneath that deeper understanding of our lawyers deeper understanding of what they want to do deeper understanding who they are and therefore enabling them to do their job quickly that has to keep going we have to keep remembering and learning as they evolve i don't think that'll ever stop but what we're going to do is make their journey on our platform as simple and easy as possible.

35:57Greg Dickason: So it's getting that balance right. We can't have 100 buttons because we've had 100 great customer sessions and each one's given us a new insight because obviously that just overwhelms you. What we've got to do is have the agent understand what you want and based on our sessions, be able to give you exactly what you need without you feeling like, oh, wow, there's something new every time I log on to LN. I'm sure you found that with some of like Claude, for example. Every time you log on, there's an update. It's changing all the time. We don't necessarily want to go there. We do think we want to change as often as they do, but we don't make that something that's invisible to the user.

36:35Yeah. Yeah, I mean, it seems like a lot of what you've been talking about are very feature-based or function-based, workflow-based in some way. So you can see that playing out in a number of ways over time. I guess one of the other, maybe one other question, I'll let you go. But, you know, I'm somebody who grew up thinking of LexisNexis as a platform that I conducted searches on. And obviously, that's been evolving dramatically over the next couple of years. But does it, do we kind of get to a point where LexisNexis is really more a platform that lawyers are kind of assigning work to and then sitting back and waiting for the end product to get delivered?

37:16Greg Dickason: I think potentially we are. There's always going to be that trust layer. We see ourselves primarily about how do we make sure you've got trust in the content that we provided in the grounding that you've got. So, yes, you can assign work to it. It does the work, but it also gives you back the trust signal. So you can click through and see, did Lexis get it right? Did Lexis Plus with Projet get it right? Because I can track those all the way through. It gives me trust in the final work output. But, yes, more and more I see our platform helping you to do your work. Yeah. Obviously, the trust aspect is a big one.

37:49And I assume as you're referring to that, you're kind of distinguishing yourself from the frontier models out there, other models out there that aren't grounded in legal content. Is that really kind of the message?

38:08Greg Dickason: That's exactly right. Because it's not just the content, it's all the work we've done in linking that, all the editorial work we've done in putting that together, creating a knowledge graph, you know, what we call a point of law graph, having a really good, robust search system. And so when we surface to the model exactly what it needs to answer your problem and your question, and that's really our claim to fame, right? The ability to give you the best answer when you ask it. And that needs to surface regardless of whether it's you asking, what's the agent asking, and regardless of whether it just wants the answer back or whether it wants to turn that into a draft or some other longer piece of workflow.

38:42Greg Dickason: That for us is that core layer, which means that everything on top of it that happens, you can trust. Greg, was there anything else that you would like people to know about the Innovation Lab or anything else we've talked about today that we haven't had a chance to talk about yet? I'd just say I'd love for people to get involved. The lab is one about us getting better product for our customers, for our market, and we're really excited about that. But the other side is if you want to build with us, we're very open to that as well. So we'd love to spend time with customers who are looking to build their agentic solutions, and we're very keen to be involved in that as well.

39:22Which is probably the most important question I didn't ask, which is how do they get involved?

39:27Greg Dickason: I don't actually have a URL. Is there a URL out there? There should be, yes. And I'll make sure the team sends you one, Bob. But also, I'd say reach out to your sales rep and they can put you in touch with me or reach out directly to me and I'd be very keen to get you involved. All right. Good. I hope we get to lose with great ideas. Well, thanks, Greg. This is really interesting. Interested to hear how this evolves and look forward to updates as it goes forward. So I appreciate your time today. That's brilliant. Thanks, Bob.

40:03Thanks for joining us for today's show. I hope you enjoyed it. If you'd like to share your own thoughts or comments, please do so by messaging us on LinkedIn or other social media, or just email me directly, ambroji at gmail.com. If you're a fan of Law Next, please leave us a review wherever you get your podcasts. Law Next is a production of Law Next Media. I'm your host, Bob Ambroji. I hope you'll join us again next time for another episode of Law Next.

40:33Thank you.