A Brief History of Everything DMS (ft Mark Denner)

Take a deep dive into the evolution of document management systems in law firms with hosts Damien Riehl and Horace Wu. Featuring expert insights from Mark Denner of Fireman & Co., this episode explores the history, current landscape, and future of DMS. Learn why systems like iManage and NetDocuments dominate the market, how AI could be poised to revolutionize DMS, and the unique challenges and innovations in the legal industry. Don't miss this comprehensive look into the essential tools that law firms use to manage their vast data securely and efficiently.

00:00 Recent Travels
01:08 The Gen AI Landscape in Law Firms
01:42 Introducing Mark Denner: Document Management Systems Expert
02:32 The Evolution of Document Management Systems
05:43 Challenges and Solutions in DMS Adoption
13:06 The Rise of Cloud-Based DMS
21:07 The Future of DMS and AI Integration
36:55 Closing Thoughts

Transcript

0:01

Damien Riehl, it's been a while. How are you my friend? I couldn't be better. How are you, Horace? I'm doing very well. So we've sort of inadvertently taken a holiday here, and you've been away. I've been busy. Tell us what you've been doing the last couple of weeks. Last week I was in AALL, the American Association of Law Libraries. We were able to talk about nerdy researchy things. The week before I was in Germany. My mom had never been outside the continental US. We said, our family farmed here in the 1500s, here in the 1600s, so that was quite lovely. Before that I was in London. So yeah, it's been a lot of traveling, but sadly less Further Comments, and maybe "lesser comments", "fewer comments". But now we're back to "Further Comments". How's your last few weeks been? It is, it's been busy. It's, uh, kids and work. Kids and work. We've been head down, bum up, trying to get our, one of our tools ready for self-service SaaS launch. And that's gonna be coming in a couple of weeks. Really exciting stuff, which I'll tell you about maybe next episode when we're talking to Joe Green and Stephanie Goutos from Gunderson. I'll tell you what I've been thinking about in the last couple of weeks, because we haven't really been talking. I've been banking up all my thoughts. And my wife knows when I bank up my thoughts, it's not a happy thing to do.

And one of them was: 1:08

this Gen AI landscape that we are in. Like right now, the two players that are dominating are Harvey and Legora. It's on the lips of every major law firm, every large Fortune 500 company. I don't know why they're dominating. I mean, I, I know why they're dominating. It's marketing. Yes, jinx. But thought, why aren't the document management systems doing something about it? Because they're the perfectly placed entities in this ecosystem to really take the lead on this. And so to help us explore this topic today, we've invited an expert on the topic, and that is Mr. Mark Denner. Mr. Mark Denner, welcome! Thanks. So good to be joining you on this. Really thrilled to have you, Mark. As Horace and I we're bouncing around ideas of people to be able to speak knowledgeably, we thought, no person better than somebody from Fireman who's really had a deep into the weeds experience for document management systems. So, really thrilled to have you on, Mark. Yeah, so for those who don't know, Mark leads the, the practice group for the document management system team at Fireman. He's got decades of experience working with AmLaw 200 firms and helping clients with complex integrations for information governance, risk document management, et cetera, et cetera. So like, while I imagine I have some level of knowledge, Mark actually has knowledge. He's been working on this and he knows this stuff forwards and backwards. And so I thought what we can do is let's go way, way, way back to history and explore why have document management systems come up in the first place? Why are they so ubiquitous in law firms? I'll start with the story of when I started practicing, I was a, a little junior law clerk working in solo practitioner's office in Manly in Sydney. And he would take off every afternoon to go surfing at 4:00 PM. It was wonderful. And he had a three room office for his solo practice. And one of the rooms was files, lots and lots of files with like colored tabs, names, et cetera. And that was our document management system. And then the next year I started working at a large law firm. Back then it was Mallesons Stephen Jaques, and we jumped into document management systems. And this, this was 2004. Why, why that huge difference between people managing all their documents in paper, and a large firm having a really sophisticated document management system. Mark, take us back to the beginning. Yeah, sure. Maybe we can start think about what actually is a law firm document management system. The world is certainly full of document management systems outside of legal, although we may not think of them very often. You've got things like, OpenText and LiveLink, stalwarts like Documentum and FileNet. Those have a pretty significant presence in other industries. They occasionally pop up at large law firms, but over the years, most of those implementations have gotten replaced by more conventional law firm DMS. And, when I think about what is a law firm DMS, I think. I mean somewhat, some somewhat self reflexively, I'd say DMS used and marketed to and sold to law firms. But the real differentiator, I guess is a concept of client matters, which are, for better or worse, the coin of the realm at law firms. But the other one is that, I always say corporate DMSes capture some of the documents for some of the users, and law firm DMSes capture all of the documents for all of the users, or at least tries to. So it's a, it's a real difference in perspective and scope. The other thing that legal DMSes have evolved to do is do a better job of capturing correspondence. So emails generally speaking these days. Those are sort of the key differences between the industry specific document management systems. So just to go to the pre-history, I myself entered the, legal technology building way back in 1997. And that was actually kind of a really interesting time. Law firms used things like Wang, and, Multimate through the 80s and, and into the 90s in many cases. But by the time I got there, that landscape had changed. And in funny ways, it was a two player industry in 1997 when I got there. So when I first started working in the legal industry, it was Soft Solutions, which was almost universally referred to as SoftSol and PC DOCS, which was generally just referred to as DOCS. They were the dominant players across the AmLaw 200 in 1997. I really like Mark's history, and when he said that firms "use" document management systems, did they? Uh, you know, I think the biggest enemy of document management system is the desktop. That is saving to your hard drive. Of course, if you don't save it to your hard drive, it's uploaded as "final", and then you have a "final final", and then you have a "final, final, final." Yes, I think firms have tried to "use" DMSes, with varying success and failure. I think coders kind of have solved this problem too. Linus Torvalds, who made Linux, he created GitHub for this version control problem. To be able to say, you know, coders are saving things to their hard drives. They're not actually uploading it in a common repository. So he said, let's create a version control mechanism to be able to say, I'm going to take this document, make improvements to it and then push it back into the master document to be able to have everyone literally be on the same page. I think that's been the goal of DMSes within law firms. And I think we've largely, with varying degrees failed at that goal. Mark, thoughts on the enemy being the desktop and ways to be able to solve that enemy. So it's funny, I regularly work with corporate legal departments who don't have document management systems. And when you talk to the lawyers, regardless of how much they might have complained about the document management system when they were at the law firm, they passionately miss it in most cases. And you're right, Damien, and I think one of the things they really miss is the version control, because that's such a, disaster on a file system if you just try to tackle it with the tools that are readily available there. And yeah, law firms definitely struggle at the individual and occasionally at the practice level with adoption. You know, we talk about in sort of social media technology, the network effect, which is sort of says that the more people who use a particular platform, the more valuable it becomes. And that applies in spades to a document management system at a law firm. And so if people opt out, you get this cycle where, where if everybody, you know, in your corner of the office building is using the system, you're more likely to use it yourself. My last joke on that is, you know, people talk about lack of adoption in document management systems, and all I have to do is say we have to take your document management system offline for the weekend. And you get the real story, which is, oh my God, it can't go, it can never go offline.

Can we do it between, you know, 4: 7:48

00 AM and 6:00 AM or some window there. There's certainly mixed feelings about document management systems. I actually think, the platforms have evolved and become more consumer-like in terms of their interfaces and usability. There's an awareness of that and evolution to try to address some of the lack of adoption. My practice history as a litigator, DMS usage was actually quite high during the drafting phase, but then there's a last mile problem where the assistant would then file it with the court. And often that filed version was on the assistant's hard drive, and then they move on to the next thing. And so actually, the true "final" version never made its way back into the DMS. And that's kind of a big problem. There's a joke that, uh, really is not a joke that we at Docket Alarm sell law firm's data back to them. Uh, because since the, since that final version is on the assistant's Hard Drive, we have the truly "final" version that was filed with the court. Yeah. Yeah. give them that "final, final, final" version." Really use me please" version. That's fair. And, and leakage is just this ongoing issue within the system as soon as those documents get out of there, the document management system becomes less valuable. Yeah, I mean, it's the same for transactional, right? Like the moment the documents are signed, if you're lucky, you used to get 'em scanned in and there's a "deal bible", and maybe they'll be put in the right place, but usually it's not. And once that happens, you hand it off to the client, and the transactional lawyer's like, "well, my job is done." And, and the client rings you six months later, they go, do you know what happened here? And you're like, "no, they, they're your documents now. Aren't they sitting in your office?" Yeah. I love that — I've never heard "deal bible" before, but I think that 11th commandment is thou shalt save the final version to the "deal bible", right? That's one of the other concepts that I try to drive with law firms, which is client service, right? Like, and, and what Horace describes is, is absolutely true. And if you don't have the final documents— and probably some of the more significant drafts, easy at hand when a client calls you back in, in five years, or god forbid, 10 years when the lawyer who did the work has departed the firm. You're not providing good client service. Which I think every law firm, wants to improve their client service. Mm, so, so back in the 90s, you Yeah, yeah. Back, let's go back to the 90s. Back to the 90s. You already had two major players. SoftSol and DOCS, and law firms are using them. You're talking about AmLaw 200, having these systems in place. What made them go away? Why do we have iManage and NetDocs today? So this is where I get to say I was there, right? Like all that stuff that went before, I wasn't present for, but I entered the market at a really interesting time. I entered the market in 97, Windows NT, I'm gonna go way back here. Windows NT and, SQL Server 6.5 came out in 1996 and they were really the first networked, Windows based, Microsoft based, platforms that you could do real client server architecture on. And that happened in close succession with Y2K and certainly for SoftSol. Soft Solutions got sold first to WordPerfect, and then WordPerfect got sold to Novell. And Novell just decided that Soft Solutions couldn't be made Y2K compliant. And the entire market had to shift on fairly short notice, and iManage sort of fortuitously appeared on the horizon just at that point, as, as things so often happen, right? At the right place at the right time. So there's a huge surge of adoption as firms exited Soft Solutions and into early days of iManage. And, and around that time, PC DOCS got sold, first to Hummingbird, and then later on to OpenText. And candidly, that just took the wind out of the sails on PC DOCS, DOCS Open. And it really failed to innovate. Failed to sell into the Soft Solutions market. So iManage just ate their lunch through the late 1990s and early 2000s and became the dominant document management system at that point. They really just benefited from being in the exact right place at the exact right time. Yeah, iManage is now a dominant player, at least in the top tier firms, but they had their own little rollercoaster history. And I wasn't there for this, but I do remember the very end of it. Mark do you Yeah. the tumultuous times? Very much so. So, iManage, um, was acquired in 2003 by Interwoven. And Interwoven got acquired by Autonomy in 2009. And then HP acquisition of Autonomy, which could only be kindly described as a complete disaster, happened in 2011. HP, the next year had to write down something like $11 billion in goodwill, of which $5 billion, they attributed to straight up accounting fraud. So, iManage certainly, and Autonomy certainly, entered into the HP world with a bang, but not a good one. I would say those were, those were tough years to be working as a partner, with iManage because there just wasn't a lot of focus. And HP, I think, didn't invest in iManage. Those were difficult years with iManage. But they, they eventually got bought back by the management team, right? This was Yeah, yeah, yeah. 2015, 2016, Yeah. ish. It was either 2015 or 2016, I'm not sure which. The original founders bought out the company from HP, thank God, and became an independent software company. And just resumed where they left off: with a real focus on innovation and doing great things. We can talk about that. But, while that was happening, we had the rise of NetDocuments. I wasn't really aware of them until 2002 or later, because they became much more popular after 9/11, when I think a lot of certainly New York City law firms realized they needed to up their disaster recovery game. So NetDocuments was founded in 1999, and they've always been a cloud-based platform. And I know in the legal industry, people like to complain and moan about the lack of being cutting edge in technology. But if you actually look back in the history, NetDocs came out at the same time that Salesforce and NetSuite came out. So before... we now call it cloud computing, but nobody called it cloud computing back in 1999 or 2000. But we've had this important vendor in our law firm ecosystem in our legal ecosystem that's been cloud since the very, very early days. I think that's a great tribute to the vision that NetDocuments had back then. Which has kind of been their competitive advantage: that they are cloud native, whereas, other vendors have been on-prem focused, and I think lawyers — of course, especially pre pandemic — were saying, "you're gonna take my on-prem server out of my cold dead hands." Uh, Yeah. Yeah, sure. So really this upstart that is cloud native, much like Clio, you know, it's weird that Clio was cloud native at the time."Nobody's gonna, nobody's gonna trust their data in the cloud, right?""Especially not my documents, right? I'm not gonna put my documents in the cloud. Are you crazy? What if a hacker gets in there?" But I think that the industry finally realized like, oh wait, I'm paying a high school graduate to guard a server, that is in an unlocked room. Maybe the PhDs in security at AWS and others maybe have better security than my high school graduate that has an unlocked server. I think that the on-prem versus cloud was a really big thing back in the day. Oh, it was huge and, and I think we should give NetDocuments full credit for being an agent of change in the industry. They, probably uniquely, changed the industry's perspective on on-prem versus cloud. Because you're totally right. In 2005, 2015 even, there was a significant fundamental resistance to going "to the cloud." And I think NetDocs had a big part of changing the industry's perspective. So interestingly, I, I think the histories set up NetDocs and iManage as having kind of competing philosophies when it comes to their technology and their design approaches. Um, maybe through circumstance, maybe through design. And I know you took a little break, and you walked away from doing DMS work for a little bit, but then you came back to it in the 2010s, and that's when you started leading the practice at, at Fireman, right? So when you came back and you saw kind of the shape of the market, who was dominating at that point? This was like 2020. Yeah, 2020 and, and even through today, it is definitely, it's a duopoly. Uh, the AmLaw 200 is either iManage or NetDocuments. That bottom 25 of the AmLaw 200 sort of cycles in and out, so it's a little harder to to track that. But, fundamentally, every law firm in the AmLaw 200 is either using iManage or NetDocuments. But iManage is still quite dominant in the AmLaw 200. So it's definitely 80%, solidly, 75% to 80% of the AmLaw 200. You do have some really prestigious firms like Skadden and, and Sullivan and Cromwell, who are on NetDocuments. But the vast majority of them, that 75 to 80% iManage in the US and an even more sort of dominant position in the UK and global markets. So it's probably closer to 95% of the 100 largest UK based firms are on iManage versus NetDocuments. But it is at this point, for all intents and purposes, a duopoly in the industry. Who, who else is playing in this space? And let's, let's just round this off because I, I know OpenText still had Skaddens even a couple of years ago, and then they moved on. Yeah. And there are a couple of outliers there. Skadden, I think is, either in the process of, or, or has moved off of OpenText. And there are, I can think of one or two other, AmLaw 200 that are in the process of moving off of a PC DOCS, OpenText, or whatever we want to call it. And you know, there's always been sort of some entries into the marketplace outside of those two. None of them's really gotten a meaningful toehold in the market. There's always been SharePoint out there. I think law firms are attracted to it as a free platform, and also a platform that has a really rich set of capabilities, but maybe not the set of capabilities that law firms need. But some third parties like Repstor, Epona, MacroView who are out there providing what are essentially custom interfaces to SharePoint, to sort of fill in the gap between SharePoint functionality and what law firms actually need. And you periodically see those, a lot of the times they enter the AmLaw 200 through acquisition. So an AmLaw law firm acquire some much smaller firm, but they bring with them a SharePoint based DMS or some other technologies, maybe like WorldDox . And then you've got some other players who occasionally flirt with the market. And one I've seen a few times in different contexts is M-Files, which is a really fascinating, different take on document management. They, I think were originally Finnish based, although I think they're now based in the United States, and so I periodically see them being considered. Outside the AmLaw 200, although you see plenty of NetDocs, and I definitely say NetDocs holds a significant part of the market in the mid and small law firm space, where they have a document management system, it is quite often NetDocuments. But you've also got the all in one platforms like ProLaw and Peppermint, which is quite popular, that have a DM capability within the platform. We talked about OpenText, we talked about SharePoint, there's a real question as to what really makes a "DMS"? Because is SharePoint a DMS? SharePoint is kind of a bucket where you put files, maybe it's a document, maybe it's an image, maybe it's a video. Right? ILTACON last year had a really good session led up by my colleague Ed Walters, where they said, what is the future of DMSes, if this is really just a bucket? So really the question is what does a DMS add on top of the bucket of files? One of which is security. You know, lawyers really need ethical walls. I have an ethical conflict with Horace's clients, so I, I can't see Horace's thing. So ethical wall is one.

Version control: 19:18

I need to check out a document and check in a document. You know, there are a lot of these buckets just don't have checking in, checking out, and changes between those two. So version control, I think is probably number two. Metadata is maybe number three. You know, which matters does this belong to? Which client? What is the type of matter, et cetera. That I think is maybe a competitive moat, between just plain old vanilla buckets and DMSes. But Mark, am I, am I missing others? No, I think that's right. SharePoint has some scalability challenges. Legal users, although there is the ability to do versioning in SharePoint, it tends not to be the way that lawyers want it to be. Some of that might be because many of them came up using either iManage or NetDocuments. So they have a very clear vision on how versioning works. And SharePoint doesn't really meet that. SharePoint has some search usability challenges that get overlooked a lot until you actually put it in play and people use it, and the search doesn't meet some of those requirements. And actually, just to go back to one point you made: architecturally, for reasons that would bore everybody to tears, SharePoint can't really do ethical walls easily. There is no such thing as an"exclusion" for access in SharePoint. In Silicon Valley, information needs to be free, man. Yeah, yeah, yeah. I guess, I guess so. So if you do want to, if you do want to exclude Horace from a particular client or matter, it's technologically quite challenging to do that in SharePoint. There are definitely a lot of ways in which SharePoint is close, but doesn't quite hit the mark on law firm document management. And I think you've seen that. Several of the Magic Circle firms in the UK invested a lot of money and time to make SharePoint a DMS, and I think they've all exited that platform at this point. They've all gone to iManage. Not for lack of trying. Yeah, that's right. Yeah. I wonder, do large language models change that calculus? Yeah. Maybe when law firms think of large language models, a lot of them think of Microsoft and Azure for that, and certainly SharePoint, fits into that architecture really well, and Microsoft is trying to really build in a lot of large language models into SharePoint. But so have the document management vendors, right? Like iManage and NetDocs both for the most part, relied on Microsoft based AI technologies. So it's competing "like with like" in that respect. Do you remember last year at ILTACON, there was a big sign right outside the vendor hall for Box.com. And it looked like for a second they were... Mark is shaking his head. I thought they were gonna enter the market. Yeah, and they've certainly toyed with the market. I think Box has been out there trying to slip in the unlocked back door, if you will, of file sharing, which is a challenge with the two legal vendors and say, sharing files in that document management system outside the law firm is prohibitively difficult still, for both platforms after all these years. And so Box periodically tries to jump in the gap. But I guess one thing we actually didn't talk about , but a 3,000, 4,000, 5,000 person law firm is a big law firm. But if you go outside of legal, 5,000 people is absolutely nothing. Um, I think a lot of those vendors look at the legal market and in aggregate and say, oh, cool, there's lots of money there. But then you do one level deeper and you're like, oh, but there's like 8,000 buyers there that I'd have to sort of go out to. So I think Box and some of those other vendors periodically flirt with the industry, but I don't think they're out there trying to displace the document management system. And I also think the legal industry's just a hard nut to crack. When I talked with a large, massive player that everyone knows the name of that, I'm not gonna say, a decade ago, they said if Google wanted to eat our lunch, they could in a second, but the legal market was just far too small. Uh, Yeah, Google's thinking in the, you know, in the hundreds of billions and not in the, you know, the, the small, tiny legal marketplace. And so the question is, yeah, "Do the Boxes of the world, do they really want to add the ethical wall functionality and that kind of thing? To be able to enter what is really a tiny market comparatively?" Yeah, I think that's probably one of the firewalls that prevents non-legal technologies from really making entry into the law firm world. I'm sure if we put our heads together, we could come up with some large vendors who've made significant progress into the legal marketplace. But honestly, most of the legal industry software players came up through the legal industry. They didn't enter the legal industry laterally. It's kind of an interesting thought as to what kind of a closed marketplace the legal industry really presents. So that, that brings me to the initial thought that gave rise to this topic in the first place. Which is given that NetDocs and iManage hold pretty much the wealth of information inside of law firms, all of their documents and emails and so on, why haven't they leaned in more to Gen AI? Why haven't they been leading the market in releasing products, in releasing features that leverage Gen AI and the documents that they're already holding? Mark, what, what are your thoughts on that? Yeah, it's really great question. You've sort of got this classic disruptor model of smaller, more nimble companies that see the legal marketplace. Understand, and let's be clear, I, I think of any industry out there, law firms are most likely to benefit from AI just given how AI works and given what law firms do. I'm sure if either vendor is watching this, they'll be annoyed with me, but, but they're, they're focused on many, many things. And AI is one of them. Therefore it's been a little bit slower to enter their space. But they've certainly made really big inroads. We can talk about sort of which document management system is better, and I usually have little patience for that particular conversation.' Because at the end of the day, they both provide very, very similar capabilities. But I think one thing that makes NetDocs unique and distinct from iManage is that for throughout its entire history, NetDocs has had one platform. It's always been the same version. Everybody's using the exact same platform. Their cost of ownership and their ability to innovate just within that platform has been much better than iManage's. iManage has got 4 or 5,000 clients. Many or most of them still probably on-premises. They're supporting dozens, like literally dozens of on-prem versions. They've had multiple iterations of their own cloud. So NetDocs I think jumped into that a little bit more quickly. Just given some platform advantages. So as of ILTA last year, their gen AI platform, which I think generally falls into this ndMAX, like a lot of software vendors, the marketing sort of outpaces a human being's ability to understand the marketing. But I generally think of ndMAX as their offering. And they were live and available for purchase, essentially ILTACON last year, and I think we'll see iManage, with Ask iManage, it's been around for over a year. But it is still mostly out there, in pilot beta, format of some part, in some sense, depending on how you think about it. They're definitely getting there. I think once they really get going with their AI products, they will have an opportunity to really dominate the industry simply by virtue of holding the documents, right? Like so, What documents get into AI, how they're secured, how client requirements are met in AI. And it's really hard when you've got a bunch of users potentially, I'm, I'm exaggerating for effect here, uploading confidential or or otherwise sensitive documents into various AI products legitimately gives a lot of law firms heartburn. iManage and NetDocs by keeping the documents that are being run through whatever model within the system, inherently have an advantage. They're right there to enforce ethical walls and other concepts of security. So there's less concern of leakage and exfiltration of those documents, intentional or otherwise. Although I agree a lot of AI upstarts got way ahead in terms of marketing, as we were talking earlier. Then iManage, I think, as they make AI their focus, I think they have a really great opportunity to dominate the marketplace. I'm familiar with both ndMAX and Ask iManage, and they are compelling, interesting, useful, technologies within those platforms that are neatly integrated into those platforms. Three questions as a follow up. One is thinking about data as oil, you know, we've talked about this a lot in this podcast. And of course DMSes have a ton of private oil, contracts the settlement agreements, all of the internal, you know, memos, that kinda stuff. So that private oil, number one, and it's a massive data set to be able to mine. Number two, what you said about iManage needing to deal with on-prem and others and, you know, a bunch of different versions reminds me of Microsoft. I follow one of the engineers of Windows, he's been an engineer since the 1980s. And he said, yeah, most of the things you hate about Windows is the fact that they have to support something from 1985 with Windows 3.1, that some mission critical government system still runs on Windows 3.1, so they have to make the newest version of Windows compatible with something from 1985. Compare that with Apple, where they just have, you know, they have one version of the OS. They go back to as far as they can go, and then they decommission, then that's that. Right? So really that backward compatibility is not just for operating systems, but also for document management systems. You're only as good as your least common denominator. And if you haven't updated in 20 years, I guess you're stuck with the old thing. That's the thing number two.

Thing number three is more of a question: 28:45

the search layer is, is really important. Trying to find something in your DMS is really important. I wonder if, Mark, you could talk about the emergence of companies like DeepJudge that are, at least offering a way to better search your DMS in a way that's maybe different than the two big players. Thoughts on on, being able to be a search layer on top of the DMS? Yeah, it's super compelling. And there's always been that, going back, literally to early days, there have always been competitors like Recommind, and RAVN, which now part of iManage, but prior to being part of iManage, a lot of their business was making document management search better. DeepJudge is part of that long heritage. And a lot of my clients are really avidly interested in DeepJudge. I don't really know where that'll head. I think one fundamental difference is that DeepJudge has some notion of training its AI model on the whole corpus of documents in the document management system. For the most part, NetDocs and iManage have focused on submitting a set of documents to AI for analysis and not trying to leverage the entire group wisdom, if you will, that's embedded in the document management system. It'll be really interesting to see how that plays out between the approaches of DeepJudge and iManage. Because again, partially to address law firms' concerns around things like security and ethical walls. Both platforms at this time very much depend on me selecting these 20 contracts and asking questions about these 20 contracts. Not saying "based on our entire body of contractual work product, what are some options for evergreen provisions in a contract" or whatever you might be asking for. I think it's interesting, like my opinion on this one, and, and it's informed by a lot of conversations with top tier law firms, is that the work products in the system belong to the client. And so if the client says, you can't train on this, then, you would have a very difficult time carving out certain files, certain clients from this training corpus that you would give to an AI. So while I find the efforts to finetune a model to be very admirable or train a model, I'm not sure what DeepJudge is doing, I think they're gonna face some practical difficulties. And while I love the DeepJudge team, the thought that comes to my mind is BloombergGPT that was created way back when, which I think cost like $8, $10 million to create. And then, like, two months after it was released, GPT 4, whatever came out. And it's like, well that's it. That, that's, that's, that's no longer cutting edge. So I think there's a, there's a great risk there. I think there's a world where obviously you have a document management system. You maybe have something like Insight+ from iManage, which is their search platform. But you also have DeepJudge out there, each providing some set of functionality that's relevant for a specific group of users, much like you have for a long time had WestKM and LSA targeting different discreet sets of functionality that are relevant for either transactional or litigation lawyers. So, maybe the future is a more diverse future rather than just, you know, a couple of colossuses. Thinking about the training on the datasets, of course, everyone was freaked out by Samsung's apparent trade secrets making their way outta the backend. And I, I think that was a shot heard around the world to say, okay, law firms are saying, "don't train on my data'cause I don't wanna be a Samsung". But when we think about how the large language models are trained, the weights favor things that are "most common" and they don't favor the "outliers." And so when you think about really what the training is, is boilerplate: stuff that is really not trade secrets, but is really just, you know, the most common boilerplate, you know, preambles of a contract. Or the most common summary judgment standards, right? These are not things that are client specific, but they're just things that we do every day, that the training is gonna be more favored towards. So I wonder if, just like the copyright lawsuits that they say, you know, it's all fine and train on something until you start regurgitating Harry Potter. It's only with the regurgitating of Harry Potter where you get into trouble with the copyright. So I wonder if it's the same idea that, the reason that clients go to law firms is for their breadth of experience. And I wonder if prohibiting the training on that breadth of experience is cutting off their nose to spite their face? That if proper guardrails are to say,"I don't want my client data to be regurgitated into my competitor's data." Yes, of course, keep that regurgitation out, but really don't you want your firm to be able to benefit from the breadth? And isn't prohibiting training, really cutting off your nose to spite your face? It's funny you say that because I've spent a part of my career dealing with outside counsel guidelines and it's very common in outside counsel guidelines to say you can't make my materials accessible for reuse, but we expect you to reuse all of your other clients' materials to produce work product for us. I mean, obviously, generative AI has brought that front and center, but that's been out there as one of these sort of, almost ironic provisions in outside council guidelines for quite some time. The fair use of materials from generative AI, incredibly useful, but I think what firms worry about and what everybody should worry about is misuse of, and, trying to get your generative AI to divulge some of the materials, the Harry Potter text, if you will. Yeah. You know, avoiding that is, is not too hard of a technical problem. But I think once it happens, that is a blight on the system. And much like our Avianca case and poor Steve Schwartz that essentially ruined it for everybody, you know, just need one or two of those and then, all of a sudden that ruins it for everybody. Knowing we're coming short on time, I wanted to share something that is probably gonna give our watchers PTSD. Here is something that, for any talk of DMSes, has to go to naming." I'm not sure which of these files is the final draft," but whether it's "final draft" or "final draft 1," "final draft 1 - send this one" "No, actually send this one.""This draft is correct.""This one has three finals.""Now it's back to only one final, but it's in caps." Then you go back to "final draft 2". This is something that anyone who has ever worked in any law firm or any organization, probably will bring them PTSD. Yeah, definitely. And you see chaos like that regularly out on file systems, and you can, if you want to recreate some of that chaos in a formal document management system. But I do think a single document with multiple versions, as most legal DMS is presented, is the gold standard. I talked to a, a large in-house, a massive Fortune 100 company. And I said, what's your DMS? And they said, oh, we just use Google Drive. And I thought this is a massive company that the entire, not just the legal department, but everybody is just using Google Drive. So Um. it reminds me of something that I think Alex Smith said to me. It's, it him or, or Alex Hamilton. And it was, all curation becomes search until all search becomes curation, or words to that, to that effect. And as you talk about Google Drive, that, that's what comes to my mind. But Google's lucky 'cause they've got great search capabilities. And metadata too. If you want it. Yeah, especially now with, with Gen AI, you can start creating metadata just by reading the files. Which again, question why aren't the DMSes doing this? But I think all of these questions, uh, may be answered in a couple of weeks when we go to ILTACON. So, so Mark, are you expecting any interesting announcements at ILTACON from the DMSes? Yeah, they, they tend to keep those things pretty secret. So I don't always know what my partners are gonna announce at ILTACON, which can sometimes be a little bit frustrating, but I get it. I will have to see exactly what gets announced there. But I do, I just maybe as my parting. Comment on this overall topic. I think historically, and Damien alluded to this early in this conversation, I think DMSes were really viewed as a necessary evil. You know, we've gotta have someplace where we can consistently store stuff and secure things. But it didn't really make the lawyers' day-to-day life much easier. And I really do think the integration of AI into the document management system completely changes that. And gives us suddenly a really compelling reason for lawyers to be in there working with the documents and the document management system. Instead of just it being a place to stuff the document. Maybe we'll retrieve it later. Maybe we won't. The combination of AI and document management systems is a profound challenge. Yeah, I, I agree. That's a, that's a good note of optimism. I could talk about DMSes all day, as I know all three of us could. And of course, if you're a listener of this podcast, who wouldn't wanna talk about DMSes for all day. But sadly, we have to come to a close. And Horace and I like to close with a bit of optimism. I, I think you've shared a bit of optimism. I, I'm a bit optimistic that large language models will be driving DMSes to do things better, faster, stronger: to do things in a way that hasn't been possible before. So that's my optimism for today that DMSes might be able to do cool things. Horace, what's yours? Oh, I'm optimistic there are gonna be gaps, and that startups and players are gonna go in and fill these gaps. Excellent. I like that. I did. Mark, what are you optimistic about? I tend to be pretty cynical about technology in a lot of cases, but I really do think that integrating AI into the document management systems will make them much more compelling systems. And, and like I said, that system effect of driving usage just makes the system increasingly more useful for everybody. So, perhaps in five years, everybody will be holding hands and saving all their documents into the document management system and talking about what a tremendous asset it is. So much fun having you on. I'm really grateful for your time and Horace, always good talking with you. And thanks to our listeners for, for listening. Until next time, listeners, and hopefully next time we won't be a month. Thanks everyone.