The Future of eDiscovery: AI, Digital Evidence, and What's Next with Level Legal's Matt Mahon

Speaker A: Hello and welcome to that tech pod. We discuss all things E discovery, data privacy, cybersecurity and tech innovation. I'm Kevin Albert.

Speaker B: And I'm Laura Milstein.

Speaker A: And each week we are talking to heavy hitters in the industry to help us break down these topics. Laura, who are we talking to today?

Speaker B: Today we are talking to Matt Mann, who, who is the Vice President of Client Solutions at, uh, Level Legal. Matt draws on decades of experience in eDiscovery to help clients develop bespoke solutions to their problems. Matt regularly shares his expertise and thought leadership, speaking at industry conferences, CLE events, law firm trainings, and more. He holds several industry certifications including SEDs from ACEDs, IGP from ARMA International and and is a RelativityOne certified pro and Relativity Certified Sales Pro. Matt also serves as the president of the ACEDS Jacksonville Chapter. Matt, welcome to the show.

Speaker C: Thank you, Laura. It's great to be here. Thank you, Kevin.

Speaker B: Um, uh, I'm so happy to have you here. From what we've earlier discovered that we worked at Ricoh together at the same time. We've come a long way from rico where there was no technology at the time, tech was just starting to. Now we're in an E Discovery world. I mean, I don't think at that time people really even knew what E Discovery was.

Speaker C: Yeah, well, now we're from, um, to your point, paper to discovery to Gen AI, which I think is the, the next generation of discovery. So it all comes full circle.

Speaker B: Yeah, it's crazy. It's crazy for our listeners who haven't met you before because maybe they didn't work at RICOH when we did. Who are you, what do you do? And why did you decide to spend your life in eDiscovery?

Speaker C: That's a great question. Uh, I'm, um, the VP of Client Solutions at Level Legal. I've been in this space for, uh, maybe 18 years. Now back to your question of why be here for so long? I was a math nerd, wanted to be a lawyer, did not do well on the lsat. Uh, people that know my story did not get into law school and just stumbled my way into paper discovery as imaging was becoming more prevalent. Loved load files because it reminded me of linear algebra and that was sort of my gateway from there into eDiscovery. And my math background helps with validation, you know, statistics, load files, et cetera. So I love the ever changing industry that we're in. And even though it's a little bit crazy, it's fun and never boring.

Speaker A: So what are you mainly talking about with attorneys today, what are the conversations you're having over and over again?

Speaker C: Well, a lot of it comes to Gen AI adoption. I know it's. Everyone may be bored of hearing that, but we've really evolved from experimenting with Gen AI. Are we going to use it really? To a more disciplined adoption. All that focuses on defensibility and validation. So how can we use it effectively, both cost effectively and driving efficient results and then how can we validate it? And there's some questions with the courts, but courts really aren't necessarily asking if genai should be used, which I think is a good thing. It's just another technology assisted review if you look at it big picture. And they're really looking at more of a safeguards, especially in protective orders, of what needs to be disclosed and how the AI is being validated. So we've seen that in Morgan Jeffries conservation law and there's new case law coming out daily. And really the takeaway is that AI is permitted. But of course they're demanding really more discipline around the use of gen, the protective order, language terms, workflow, protections, documenting so you can defend it and then even deletion protocols to ensure that privacy and confidentiality are protected.

Speaker B: You were right when you said it's not boring. Here we go, getting into the nitty gritty of excitement. One of my favorite things with technology is really the trends. It's like, is this going to be here to stay or is this going to be like cal tar gen AI? Like what's real? What is even just thinking like when Brain Space first came out and they were like this is true AI and I was like, well, it is kind of more like keyword searching but it's like really aesthetically pleasing. Like you like it because it looks really cool, but is it AI? And we've just come such a long way in E discovery. I guess what I'm trying to ask here, like what trends are you seeing in in the E discovery market during the first half of 2026? Like what's exciting to you, what's concerning to you? What are you like m is that real in. You know what I mean?

Speaker C: Well, two things. Gen adoption is exciting, but the other trend that actually comes along with Genai and probably is led by Genai is this is as crazy as it sounds, a greater adoption of active learning and analytics workflows. So because it's the safer AI that is out there and we can talk a little later about some of the requirements that council is receiving from the corporate clients to Use AI, what does AI mean? It has a very broad definition. And so it's really important to ask, when you say we have an AI use mandate, what do you mean by that? And then let's ensure that we're not just checking an AI box. But whatever AI is applied, it's defensible. It complies with your client's acceptable Gen AI use policy if it is Genai, and it leads to some results, whether it's by surfacing key evidence early on, whether it's cost savings, or hopefully all of the above. But it's not just checking an AI box to say you use it. And my thoughts are it's being driven further because we talk about Cal and Tar. When Tar 1.0 came out. And even active Learning, people didn't have access to that on their work computers, on their PC and Windows, or in their cell phones. They didn't have active learning at home. Unless you're in a Relativity Reveal Casepoint or any other eDiscovery workspace, you weren't using active learning. Well, everyone's using, mostly everyone's using Genai, whether it's through Copilot, Claude, ChatGPT, et cetera. So they either have it at work and, or at home. And so they're demanding use of it in the workplace. So I think that's why we're seeing a faster adoption to Genai than we did initially to your TAR 1.0 and TAR 2.0 active learning workflows.

Speaker A: But one of the things that's kind of fascinating to me is sort of the cost of all this. AI is supposed to make us more efficient. And so you assume costs would go down, but. But it hasn't really happened that way. You know, we're not really seeing the AI is driving cost down. Why do you think that is? Why hasn't it really, you know, sort of translated into cost savings at this point?

Speaker C: It's just too soon. I hate to say that, but there's a cost associated with technology. There's also a cost associated with the expertise. I think there's a significant gap in expertise now, similar that we saw when eDiscovery first became a thing. So you can have the tool, but do you actually have someone that knows how to use it effectively and defensibly? Those are two different things. And then once you have that training and you obviously have access to the tool, I think costs will come down. I remember when we go back to the these imaging days, ocr was like 10 cents an image. I mean, how ridiculous does that sound today. So I think you all have similar uh, things with Gen AI, but there's a cost associated with it and with servers technology, I think not everyone fully understands some of the backend hard technology costs and that will though eventually and still is driving significant cost savings. So if you use it appropriately and it's not appropriate necessarily for every single case, but there's a case where it applies, I, uh, think you'll see significant cost savings. It's just how it's applied. That's where we talk about sampling full circle with my math background doing random sampling or to uh, understand what, what is in our data set and how can we effectively use what technology in a way that's going to drive defensibility and cost savings. And sometimes that just goes back to using your active learning analytics that are included in your ediscovery review platform. And sometimes it's looking at Gen AI. So it just depends on each specific case scenario and those case contours.

Speaker B: I like your answer. I want to pivot here. I want to ask you something completely random. You might be uncomfortable, so trigger warning. But if uh, you're up for the challenge, my first question is, are you a vegan vegetarian? Do you eat meat? Like, where's your stance on meat?

Speaker C: I have two pit barrel smokers and a grill. So I definitely.

Speaker B: Okay, thank God. If you were like vegetarian, I'd be like, this question isn't going to work. Okay, so you like meat? All different kinds. Like if I was like, hey, here's a brisket, here's a steak. This is cow, this is lamb. Like are you adventurous or you like only beef or only chicken or like, what's your, what's your thing?

Speaker C: Yeah, extremely adventurous. And since I started smoking on the pit barrel, you get to even experiment cuts of meat you've never cooked before. So that's been fun.

Speaker B: You are literally a dream. Let me talk to you about this, Matt. So here's what I was thinking. I like to think of ediscovery as meat. And I know that sounds really weird, like what do you mean? But for me, I love meat. It's not for everyone. Just like Ediscovery. When we're talking and you're saying it's not boring, it's not for everyone, but for those of us that are in it, we're just like, you don't get it. Like it's, there's so much to it. Even when there's nothing to it, there's so much to it. A steak isn't just a steak. I can give you the same cut as I give Kevin. And based on what you just said, I want to try your steak, not Kevin's, even though it's the same piece because I believe that you're going to cook it better. Sorry, Kevin, but it does seem that way based on this conversation. Kevin's like, yeah, that's true. But here's my thing. If I were to tell you, or actually, Kevin, I'm going to ask you a question. Kevin, if I were to say, would you try kangaroo? What would you say?

Speaker A: I think I probably have some follow up questions, but I probably would try it once.

Speaker B: If I said, would you try rhino? Would you try it?

Speaker A: I think same, yeah, sure.

Speaker B: If I said, would you try human?

Speaker A: I mean, I'm not going to eat a human. That's ridiculous.

Speaker B: Okay, so here's my question though. Here's my question. Right off the bat we say no to something. And I'm not so don't eat people. But, um, but right off the bat we have these specific boundaries. I think in E Discovery we have the same boundaries and I think it is fair and right to have the boundary of let's not eat each other, right? Let's not be cannibals. But, uh, we're already eating a cow. Some of us, we're already eating a chicken, we're eating a lamb. Why not a kangaroo? Why not some of these other meats? Right? But some people would have answered that question and been like, absolutely not. Uh, Matt, I put Kevin in because I actually thought he was going to say no to some of these. Whereas you, I thought you were going to be like, yeah, I grilled it up last night. It's really lean protein.

Speaker A: Some people are like, deer is really like, they're adorable. So you don't eat them. Some people don't eat venison.

Speaker B: Yeah. And I'm like, well, there's so many of them and they're delicious. But, but my point is, and to tie this with E Discovery, um, on this long ran, something that seems really weird to other people is great to other people. And I think with E Discovery, we get so used to the one thing which is in the meat world, we'll say beef, right? You get so used to beef that you don't even bother trying anything else that might be delicious. Pork is great, but if you have your mindset that it's not good, then you're not going to bother trying it and you're just as happy with beef. But there's limitations to that. So my question and Tying back to all of this is when we're looking at all of these different products and we're looking at all of the different things that are on the market there. And we know we love a steak, we love chicken, maybe a kangaroo, maybe an alligator, who knows? We're feeling adventurous, but we also don't want to eat humans. Where is that line with discovery? When is it like too many products but also is it not enough? And I don't even necessarily mean with products, I just mean in general when we're talking about eDiscovery and we're talking tech. How do I know when I have the perfect bundle?

Speaker C: Wow, that's a great question. How do you know the perfect bundle? I think it's case specific. It's one size fits one. You may have the perfect bundle for one matter and it may not be the perfect bundle for the next. So maybe back to the meat analogy. Maybe depend on the day of the year. I mean Thanksgiving, I could cook up a great steak, but maybe not appropriate. I have to go turkey on Thanksgiving day. So just because I have to do a turkey.

Speaker B: Yeah, you have to.

Speaker A: Yeah.

Speaker C: So I, uh, mean maybe you do a ham as well, but I don't think there is a perfect bundle. And as, as data sources continue to evolve, now we have Gen AI chat logs, for example, as we continue to have new data sources, new data types, as, uh, technology continues to evolve, I think we always strive for perfection, but I don't know if perfection exists. I think we have to continue to evolve. So what's perfect today may not be perfect tomorrow.

Speaker B: So that's perfect answer. You literally gave the answer. Because if you were to ask me what's the perfect meat, I can't say steak because I want to maybe have, you know, maybe I don't want steak tomorrow. Maybe I want tacos, maybe I want turkey tacos. You know, I was trying to differentiate them but like you can't say. And I think that is the right answer. And so to wind this all back with you here, I think it also comes down to a lot of people will be like, okay, I can't make the perfect bundle, but can I make the perfect pricing? So with that, do you think the traditional pricing models in eDiscovery still make sense in AI driven in an AI driven world? Or is the industry heading towards something different, like not just steak but dried stack? I don't know where I'm going.

Speaker C: I'll say we are absolutely going somewhere different. I know that's a level legal. We're Taking this pretty seriously with how we're looking at pricing. And as we continue go back to math. Why is everything going back to meat or math in this conversation? But it seems to going back to math you need more data points to drive new pricing models. And so I think as we continue to see more adoption of gen ah workflows, more success, we receive more data points. We can be more creative with alternative fee arrangements. Flat rate pricing per matter or flat rate pricing per stage in a matter of Absolutely. Get more creative. As more technologies evolve and the more we get better at using them, more expertise with them, more experience with them and we have more experience with those results, pricing will evolve. And I think if people don't evolve their pricing, they may be left behind.

Speaker A: I mean we talk about this with law firms a lot. Like uh, about the billable hour. Do you think the billable hour goes away or do you think it just sort of change. You think firms will end up just changing how they use the billable hour?

Speaker C: I would say it probably depends. That's probably not a lawyer. That's a lawyer answer. I think it'll depend on the matter type and what the engagement looks like. But yeah, I think there will be more alternative fee arrangements and those could take different shapes and sizes. Again those could be stage related. There could be some hourly component. But I think it also comes back to transparency and communication. Uh, we need more transparency and communication. We need more conversations as technology evolves and collaborating together with corporate council, outside counsel, with technology providers, certainly alternative legal service providers. Put everyone in the same room, collaborate and let's create a workflow and a solution that's matter specific that's going to drive the outcomes for that specific case. We just need to see more of that. And uh, that for some reason with COVID everyone going to the remote offices, people stopped talking. We just need to talk more.

Speaker B: I want to talk more. I always want to talk more over uh, like a nice uh, meat. Now that I'm just thinking about your smoker. I think I'm hungry. If y' all can tell from this. I'm definitely hungry.

Speaker C: There's nothing like a nice day cabernet and you discover you put the three together. The perfect night.

Speaker B: Should we make an event? Because that sounds nice. Matt, would we want to host it because you got the smokers.

Speaker C: We love to convene people around a table so it might be something we do.

Speaker B: I'll bring the wine. I'm in. Okay. Let's get, let's get really crazy. What kind of. What kinds of evidence do you think legal teams aren't paying enough attention to, especially as AI assistants become part of everyday work?

Speaker C: Well, the obvious answer is in some of the AI chats and we've seen some of that evidence come to bear in some recent cases, but that needs to be put front and center onto your custodial interview list in terms of what data sources are potentially relevant for a case. And if that question isn't being asked, it needs to be one of the fun ones I always bring up. Not that I would ever do anything nefarious, but if I had to communicate digitally about something potentially illegally, I would do it in some Xbox Roblox type of chat platform because people aren't collecting from Xbox or PS5 consoles.

Speaker A: Right.

Speaker C: That would be potentially where I would hide evidence. So something else to consider as people try to get creative on how they communicate and do bad things.

Speaker A: Yeah, that reminds me, I have a friend who that was how he discovered unfortunately his partner was not being faithful was in the the words with friends chat. They were like, that was where they were doing their flirting. Which I was like, wow, that is a uh, spot that I don't think anybody really thinks about. Those different, you know, like in game conversations are so funny. Yeah, interesting me. What's a belief or assumption that most people in Ediscovery hold today do you that you think will be outdated in let's call it five years from now?

Speaker C: That's a good one. Five years from now. I don't know. I, I, I almost think that a lot of assumptions aren't outdated. They evolve. I would hope that people are adopting gen use cases and technology more effectively and thinking front of mind, thinking about data sources differently and hopefully just overall collaboration. Uh, another pet peeve of mine and ours at level legal is sort of the tunnel vision associated with each stage in discovery lifecycle litigation holds put in place and then maybe someone comes in and collects the data like well my job's done. And then that data is handed off to a uh, Ediscovery analyst or processing team like oh, my job's done. And they hand it off to an eDiscovery PM to load in the relativity. Well now it's the review team's problem. This is a holistic approach. Everyone needs to be thinking about every stage and every stage is everyone's problem. So maybe that's the belief that hopefully everyone understands that with technology, with better workflows we may be able to jump from one stage to another faster, especially with more information governance maturity that they think of every stage in the process as everyone's problem versus passing the buck to somebody else. It's not the answer you're looking for,

Speaker B: but that's the answer I'm looking for. It's absolutely the answer I'm looking for. Not my problem. Well, let's, let's get into some exciting news with Level Legal. There's a really cool company when it comes to E discovery. There's so many discovery companies and we really try to wean out the ones that were like uh, this to the ones that were like oh, you guys are doing different things and you have some really exciting news to share. I know you guys secured a new ranking this year. Can you talk a little bit about that and what it means for you and your team?

Speaker C: Yeah, we were actually ranked Band three as an E discovery provider in chambers and we really are static to receive that recognition. I mean that chamber's recognition ties back to uh, earlier points about discovery being a team sport and our clients giving us the opportunity to collaborate with them, have those conversations, talk and solve those one size fits one problems and be that trusted partner. Through discovery we've been allowed to collect from some unique data sources that have had posed some challenges and clients trusted us and our expertise to do that. And again it goes back to collection, to review and production. How do we collect this in a way and one thing I'm specifically thinking about right now is it tickets and chats. How we take the ticket, apply the chat, um, versus looking at some antiquated know spreadsheet or CSV file that's hard to determine. So really thinking big picture with our clients and them trusting us to, to drive those solutions also you know we've been known in our 17, 18 year history as for, for a manager review excellence and it's really great to be recognized for our ediscovery chops. I think we've invested heavily in our ediscovery offering and uh, our ediscovery expertise. And again we're very grateful for our clients that provide this type of feedback and recognize us and for chambers to see us as Band three. So exciting news.

Speaker A: That's awesome. We uh, also heard you have an upcoming webinar with the uh, EDRM on digital evidence. Can you tell us more about it and how people should sign up?

Speaker C: Yeah, colleague, uh, Capri Miller, she's a client solutions partner here at Level Legal. She's a licensed attorney, been doing ediscovery and review even, even started as a, was a review manager uh, for years and so she spent A lot of time preparing really it's on case law. So the whole CLE is grounded on case law. So it's not theoretical. It's real live cases with challenging data sources. So some of the learnings can be applied immediately by the attendees and not just applied immediately, but then there's also some foundation and defensibility associated with it because there's case law that backs it. So I think people enjoy stories versus theories and real life stories. And so she does a great job telling the story, uh, applying the case law to the story and then also proposing some solutions that really anybody in our space, whether it's a litigation support professional to a lawyer, can take away and again apply to their business immediately.

Speaker B: You've been an excellent guest today and we're, we're fans of Level Legal. We think it's a, it's a cool company. We think you guys are doing a lot of exciting stuff. We're exciting to see what's to keep coming down the line as, as you guys keep growing and changing and bringing in new and exciting uh, technology and features which is super cool and we really appreciate you coming on today. Before we let you go, is there any advice that you would give to our listeners that maybe are interested in Ediscovery but don't know where to begin or maybe they want to get in the industry and you're like turn around, go back to rico or just any, any thoughts or advice that you would want to give anyone listening today.

Speaker C: I mean if you want to live in, or work in a fast moving industry again where there's never a dull moment, this is a good place for you. But it does come with some sacrifices and hard work. Some, some casework ebbs and flows a little bit. But be curious. I think curiosity is really an important trait for anybody in any successful business. Right? Be curious. Look at the different platforms, look at the different technology. You mentioned the uh, perfect tech earlier that we said doesn't exist but we can learn from each different provider, each technology application, get your certifications, play with the tools. I mean that, that's how I learned. I mean I'll go back to my scanning days and load files. I used to take documents and scan them with barcodes and to see what load file ah, you know would come out. So get into the tech, use the tech. You're going to have some failures, you're going to have some wins, but use those opportunities that you have to play with it so you can have that experience, see what works well, see what doesn't. Work, and then you can share that with your colleagues and clients. Have fun with it again and be curiously, my number one recommendation to anybody in the space.

Speaker B: Amazing. Thank you. So, Kevin, we just spoke to Matt Mann, the vice president of client solutions at, uh, Level Legal. Level Legal is a cool company. Ediscovery is so boring to so many, but as Matt said, not boring to him. Not boring to you. Gavin, you've been in Ediscovery for how long have you been in eDiscovery, Kevin?

Speaker A: 26 years, gross.

Speaker B: 26 years. I mean, that is older than I am. Not really. It's not. It's not. But you know what? Let's go, let's go with it because you know what the thing about age and technology is? Prove it. You know, and that's what I like about Level Eagle is they do try to prove what they have is real and what is good. And I think Matt's background definitely proves that. He could probably smoke a good steak and could help you with eDiscovery. But what about you, Kev? Tell me your true tech takeaways.

Speaker A: I mean, I thought it was a good one. Again, I like meeting new eDiscovery people. You know, again, I've been in this industry forever and so it's always kind of interesting when I come across somebody who I haven't met yet. And I like that, you know, again, you can still feel that he really cares. He really gets it. We're in a sort of an eDiscovery transition period where the, the volume and the complexity of data is changing and the way the industry is managing it is really changing. And so I think the pricing conversation was a good one. I like the, you know, the idea of kind of the way they are thinking about it and kind of being real about like the fact that maybe we will get to flat fee engagements or we will get to, you know, different things depending on sort of the shifting client expectations. And so I think the overall common theme of this, uh, episode was about adaptation. And I think he did a really good job kind of explaining that. And I think ultimately it's, uh, you know, I think we're in an interesting time and I think it's only going to be more interesting as we go farther down sort of through the AI journey. What about you? What are your tech takeaways?

Speaker B: No, absolutely, Kevin, you don't need to ask. What about me? You know how I feel about Ediscovery. And Matt Mann, I think he's great. I think Level Eagle is great. I can't say that enough. I've said it probably at least 16 times throughout, um, the episode today. I think that we need to pay more attention to some of the events that they're having. I think it's awesome to see the progress that they're making. I'm just excited to see what's next and what's to come. And if you're not paying attention to Level Legal, this is a new discovery company you should be paying attention to. If you're sleeping on this company, well, then we know you're not sleeping on a Tempur Pedic. And you know, you don't care about your, your health and your back and your life being let alone your technology or your discovery is or, uh, your reviews. So step up, get a Tempur Pedic and follow Level Legal. Get informed on Level Legal. And if you want to talk to me about mattresses, I do have a Tempur Pedic. I highly recommend it. And I do love Level Legal and highly recommend them as well. So reach out to us@contactattechpod uh.com. head on over to www.thattechpod.com. um, enter your email to subscribe. Get some swag. Do. Make right choices. Do the right things. Just feel good about yourself by following us and being our friends on social media. That's right, we're there. Head over to LinkedIn.com thattechpod. Give us a follow. We'd greatly appreciate it and anywhere else we may or may not be on social media. Kevin, is there anything else our listeners can do?

Speaker A: Yep. Go to wherever you get pods, Apple Podcasts, Spotify, iHeartMedia or what have you, and leave us a review, leave us a rating, let us know how we're doing. All that stuff helps us with the algorithms and gets us out there more, which is much appreciated. Thanks, everybody.

Speaker B: Thanks, everyone.