Ep 248: Epiq's Global Legal Solutions Leader Roger Pilc on How AI Is Transforming Legal Services

Primary Topic

This episode delves into how Epiq is integrating AI to transform legal services, enhancing client workflows and efficiencies.

Episode Summary

Roger Pilc, Global Legal Solutions Leader at Epiq, discusses the profound impact of artificial intelligence (AI) on legal services. Pilc explores the integration of AI in Epiq's workflows, focusing on legal service management, from litigation to compliance, leveraging technologies like Amazon Web Services and Azure. Key innovations include AI-driven analytics, document processing, and eDiscovery enhancements, which improve case outcomes and operational efficiency. Pilc emphasizes the blend of services and technology development within Epiq, portraying a shift from traditional methods to sophisticated, outcome-based solutions driven by AI advancements.

Main Takeaways

  1. AI Integration: Epiq has integrated AI into its services to enhance case outcomes and efficiency, shifting from traditional methods to sophisticated solutions.
  2. Technology Development: The transition towards a hybrid model of services and technology, with significant investments in AI, is a strategic move for Epiq.
  3. Client-Centric Solutions: The focus is on creating bespoke AI solutions that cater to the specific needs of law firms and corporate legal departments.
  4. Generative AI: Epiq is embracing generative AI to innovate in legal and compliance fields, foreseeing major shifts in the industry's approach to technology.
  5. Future Outlook: Epiq plans to expand its AI capabilities, further blending technology with legal expertise to provide comprehensive legal solutions.

Episode Chapters

1: Introduction to Roger Pilc

Roger Pilc discusses his role at Epiq and the company’s broad capabilities in legal service management. He highlights the integration of AI into these services to enhance client outcomes.

  • Roger Pilc: "Our focus is helping clients manage legal services better, from litigations to legal operations."

2: AI and Legal Services

Exploration of how Epiq uses AI to transform legal services, focusing on bespoke machine learning models and analytics to enhance client services and operational efficiency.

  • Roger Pilc: "We're integrating AI to create more efficient legal workflows and improve case outcomes."

3: Client Impact and AI Strategy

Discussion on how AI strategies are tailored to meet client needs and the impact of these technologies on traditional legal practices.

  • Roger Pilc: "AI is not just about technology; it’s about creating better outcomes for clients through integrated services."

4: The Future of AI in Legal Services

Roger Pilc shares insights into the future developments in AI at Epiq, emphasizing ongoing innovations and the potential changes in legal services delivery.

  • Roger Pilc: "We're continuously exploring new AI applications to stay at the forefront of the legal services industry."

Actionable Advice

  1. Embrace AI in Legal Operations: Integrate AI technologies to enhance efficiency in legal workflows and operations.
  2. Stay Informed on AI Advancements: Keep updated with the latest developments in AI to leverage new tools and services that can benefit your legal practice.
  3. Consider AI for Document Management: Use AI-driven tools for document review and management to improve accuracy and reduce time.
  4. Implement Tailored AI Solutions: Work with providers like Epiq to develop bespoke AI solutions that align with your specific legal needs.
  5. Prepare for Future AI Trends: Anticipate and prepare for the future impact of AI on the legal industry to maintain competitive advantage.

About This Episode

In recent months, Epiq, a global company providing technology-enabled legal services, has announced new artificial intelligence and analytics features built using the AI capabilities of Amazon Web Services. These new features include a framework for building, training and deploying bespoke machine learning models as secure APIs for customers; integration of Amazon Bedrock for custom copilot development using a range of commercially available large language models; and other features.

Joining LawNext today to provide details on all this and to explain what it means for Epiq’s clients is Roger Pilc, president and general manager of Global Legal Solutions at Epiq. With Epiq since 2019, Pilc is responsible for driving strategy and execution around organic growth, strategic acquisitions, product development, technology, and innovation for a broad range of products in areas including information governance, forensics, e-discovery processing and hosting, managed document review, and advanced analytics.

Pilc and host Bob Ambrogi talk about Epiq’s evolution from primarily a services company to one that also develops its own proprietary technology, while also integrating with a range of technology partners. They also discuss Epiq’s recent initiatives to use AI to further enhance its products, including the Epiq Service Cloud, Epiq Discovery, and the Epiq AI Platform. We also hear Pilc’s thoughts on how AI will impact the future of legal services and the future of Epiq.

People

Roger Pilc

Companies

Epiq

Books

  • Leave blank if none.

Guest Name(s):

  • Leave blank if none.

Content Warnings:

None

Transcript

Roger Pilse
The key is to be delivering an end to end outcome for clients, a client like a whole workflow or a service. So the fact that of our $600 million business, $500 million of it is essentially a managed service, it helps us click in this AI functionality into an end to end workflow that's providing an outcome.

Bob Ambrogi
Today on Lawnext. In recent months, Epic, a global company providing technology enabled legal services, has announced new artificial intelligence and analytics features built using the AI capabilities of Amazon Web Services.

These new features include a framework for building, training and deploying bespoke machine learning models as secure APIs for customers and integration of Amazon bedrock for custom copilot development using a range of commercially available large language models.

Joining me today to provide details on all of this and to explain what it means for Epic's clients is Roger Pilse, president and general manager of global legal solutions at the company with Epic since 2019, Pils is responsible for driving strategy and execution around organic growth, strategic acquisitions, product development, technology and innovation for a broad range of products in areas including information governance, forensics, e discovery, processing and hosting, managed document review and advanced analytics.

This is Bob Ambrogi, and you're listening to law next, the podcast that features the innovators and entrepreneurs who are driving what's next in law.

Before we get to that conversation, please take this moment to learn about the sponsors whose generosity supports this podcast.

Get back to practicing law with practice Panther. Save over 8 hours per week with advanced features such as automated billing, secure client portals and customizable workflows so you can get back to what you love most.

Join the tens of thousands of lawyers who have made practice Panther their preferred choice for efficient practice management. Visit practicepanther.com demo now to discover the difference practice Panther can make in your firm today.

Now on to today's conversation.

Roger Pilse
Roger, welcome to the podcast.

Roger Pilse
Thanks for having me, Bob.

Roger Pilse
Good to talk to you again.

Yeah, I hope you're doing well.

Roger Pilse
Yeah, for sure.

Roger Pilse
A lot of things I want to talk about during the time we have together here. But first of all, I kind of wanted to just start with some general overviews of you and what you do. Your president and general manager of Epic's global legal solutions business.

What does that mean? What's encompassed under the umbrella of that business?

Roger Pilse
Yeah, it's a $600 million business, Bob, and it's a broad set of capabilities that provide what we call legal service management to our clients. And when we say legal service management, for us, it's helping our clients manage legal services better. And it means getting to the outcomes they seek, whether it's winning and litigations, or responding to a government investigation, successfully running legal operations better, reviewing contracts more productively, succeeding in an antitrust matter with the FTC as a broad set of capabilities we do for our clients, implementing legal technology.

And so within that $600 million business, at this point, Epic legal solutions and EPIC is serving all the major sub components of a corporate legal department. So litigation group for sure, but regulatory and compliance. We were named Microsoft's compliance partner of the year twice. Commercial and contracting legal operations. We've been named number one in legal ops, consulting, M and A and IP. So at this point, we're providing people, process and technology solutions for all aspects of a corporate legal department, and that's 70% of our business and then 30% of our businesses to law firms. So we're serving all the practices in law firms and the let's support group as well, and different groups like knowledge management.

Roger Pilse
You came to the company in 20,000. In 20,000 in 2019 at least. Yeah.

I have to wonder what it was like for you. I mean, you were with the company, you probably came into the job thinking your job was going to be one thing, and then along came the pandemic and it probably had some dramatic impacts on exactly what your job was going to be. Is that true? What was that like?

Roger Pilse
Yeah, it's, you know, Bob, I was just telling a friend this morning, it's been the best job of my career, actually. I've thoroughly enjoyed the last five years and I'm not exaggerating. So the legal industry and the compliance industry are in a very, very exciting phase of transformation. And in terms of my own background, I've been in leadership roles in three different types of companies, definitely have led software and technology companies, and spent a lot of time working in AI. I've run managed services and outsourcing businesses and consulting businesses and hadn't had much of an experience in the legal industry before five years ago. But what's been really, really exciting, Bob, is the bringing of those capabilities, whether it's around analytics or mastery data management, or certainly AI these days, or professional managed services, professional outsourcing, and I've worked at McKinsey professional Consulting in this industry.

It's been the most exciting stop of my career and it's been really, really gratifying for us and for Epic to help general counsels and chief legal officers manage their corporate legal departments better and get to the outcomes they seek being successful, saving money, implementing new legal technology that wasn't available two years ago, getting investigations done quicker, getting antitrust deals approved, implementing new CLM solutions. And so it's been very gratifying to see the clients we serve being more and more successful in what they're trying to accomplish and save money when they do so. And on the law firm side as well, it's been super, super gratifying to help lit support teams, but then expand and help antitrust practices, white collar practices, help implement the latest knowledge management solutions.

All those prior years before the legal industry, working on digital transformation or data driven transformation or new business models, new managed services to be able to help clients get the best of that in the legal industry after seeing it at work in other industries has been really gratifying. So certainly COVID was certainly interesting and in some ways it accelerated everything I just talked about, experts like you full well know. So it really accelerated the digital transformation of this industry.

And many of the types of transformation that the industry was, was seeking in some ways got accelerated as we went through COVID and needed to digitally transform. So it certainly was a, was a hard time, but I think has made our company a better company, I think has helped our clients ultimately as well.

Roger Pilse
Yeah, well, I mean, a lot of litigation for a period there in 2020 anyway, and I guess into 2021, a lot of litigation just shut down and courts shut down. And I mean, I assume a lot of the background, the discovery processes and other document preparation and all of that continued on to some extent, but in a whole different way.

Roger Pilse
Right, exactly.

Roger Pilse
You touched on this when you were describing the company earlier. But it's interesting to me that Epic kind of started out as primarily a services company and has evolved into being more of a hybrid, both services and technology development company.

Roger Pilse
Right.

Roger Pilse
I mean, is that fair to say? And, you know, so where are you now? What is the balance within Epic and within illegal work that it does between providing services and providing technology to support those services?

Roger Pilse
Yeah, it's been a very neat evolution over the last ten years, and you're right, Epic's been most known over the last decade as a services provider. But at this point, we've got tens of millions of dollars of revenue in SaaS solutions for our customers, technology solutions for our customers, and close to 300 software developers and data scientists that are actually just creating technology solutions for our customers. So it's a big part of who we are and a big part of what we do. And one thing that has been a key foundation of the last five years is we built something called the Epic Service Cloud, which is one uniform service creation platform for the whole company. It's where we deliver services as well. Across all those different use cases, across all those corporate legal department subcategories and the subcategories of practices within the law firm. We built one epic service cloud as the platform for all our applications and all our technologies. So we're up to 1200 customers across our business units leveraging the epic service cloud, and we've had great success. And so within those different subcategories, we have both proprietary technology that we've built ourselves. Related to your question, and we also resell third party technology, but over time. Related to your exact question, Bob, the mix of proprietary technology has grown.

If you take processing, for example, maybe five years ago, that was mostly third parties. And in the last two quarters, over 50% of our processing is done with Epic discovery processing. We've processed as much as 100 terabytes in a single month. With epic discovery processing on the review side, that's often our own technology, epic discovery. And then I'm sure we'll talk about AI a bit, cause that's the topic of the day. But we've launched our own AI platform and have launched a broad set of genai based solutions as well. Based on our own technologies.

Roger Pilse
Yeah, it was just curious, like on the technology, I mean not on the, on the ediscovery. You have a discovery platform, you also partner with relativity and perhaps other e discovery vendors. How does that work? I mean, when does a customer likely to be using, or when is a corporate customer particularly likely to be using your technology versus a third party technology?

Roger Pilse
Yeah, for sure. So it depends on the subcategory of the use case. So from a processing standpoint, it's often transparent to the user to some degree. And so there's really no limit on the size or scale of an approach or of a matter we can take with our processing technology. Some customers may have a preference for a certain processing tool and will, will serve the customer's needs there. From a review standpoint, there's essentially no limit to our scale at this point. But I would say the main driver of a client's decision would be if ease of use and simplicity is very, very important to them, so they'll lean towards an epic discovery. For that reason. If self service is important characteristic, it might be internal investigation, communications, internal employee matters. And the neat thing is, and relativity is a really important partner of ours and we have a great relationship with them, if customers have a matter that grows to a certain size and they think it will need a certain richness of functionality. We have a really good integration with relativity and we just move the matter into relativity. And if they need a broader set of reviewers who are trained on that tool, that works seamlessly as well. So it's been great. And the good news, Bob, is everything we do is wrapped with the same epic access portal. So whether it's epic discovery or relativity or other third party or proprietary tools, clients can have the same single sign on and it's federated.

They have the same common set of reporting. So we've got a broad set of power bi based reports for all those use cases I mentioned, and it could be related to e discovery or legal spend or legal operations and understanding their different matters. So we have the ability to provide seamlessness to customers from a reporting from a self service standpoint and from a single sign on. So it makes things very, very simple for clients, which is great.

Roger Pilse
If I am a customer using you for managed services as a managed services provider, is there an advantage to me to also be using your proprietary technology, or does that not really matter with.

Roger Pilse
Your cloud platform as the managers? Freder I would say that we're one of relativity's biggest partners. I don't know if maybe we're the second biggest. So we've got tremendous expertise in both Rhel server and rel one, and we've got the same degree of analytics and reporting for both. So I don't think a client's necessarily disadvantaged if we're using something other than our own technology. We've got people trained in operations and in client services on both our proprietary and not. I would say that the advantage to the customer is in two things. I would say, first, the economics. So the economics are also frequently very superior with our own technologies. And then secondly, the ease of use. So there's just slightly better integrations between epic access and the reporting and the self service ticketing and our own technologies. But vast majority of our business is to be a great partner with third parties, so we try to make sure we're great at whatever the client wants us to use.

Roger Pilse
As you alluded to. The other big thing that's happened since you came into this job, besides the pandemic, is the eruption of generative AI and widespread use of AI and all sorts of different applications within legal.

So just generally we can drill down into it a little bit, but starting generally, how is epic deploying AI at this point?

Roger Pilse
Yeah, so it's been a fun journey for sure. And someone with your expertise in the space you're quite familiar with the amount of AI based work we've done on prior generations of AI before deep learning and generative AI. So before generative AI hit, we were already doing about 1400 classic ML based projects with clients a year as part of the epic service cloud. And we had a very large consulting organization around that, and so that hasn't abated and it's actually grown. And so with tar tar and that sort of thing.

Bob Ambrogi
Yeah.

Roger Pilse
And early case assessment, I believe our case insights team, they were up about 280% year over year in the first quarter, if you could believe that, in terms of consulting. And so it's both around litigation, but it's also around investigations. And we've seen a distinct move to the right of the EDRM and consulting around AI related to deposition prep and trial prep. So we were doing that for a while. And then when generative AI hit, because so many of us in the management team and in our development organization spent a lot of time in, in AI, including in e commerce and other areas, we really sought to embrace it quickly. And our goal is to be the leading provider in legal and compliance around the use of generative AI. And that's both as a tech company, what I just talked about, and our tech solutions that I'll get into, but both as a managed service provider and as a consulting.org. So within the last year, we ended up building a generative AI based platform in the heart of the epic service cloud. And it happens to be LM agnostic and has capabilities such as RaG and other technologies and administration around it. And then we use that platform to create a set of what we call skills on top of it that get packaged into use cases. So the skills could be summarization, they could be chatbot like skills, they could be classification, they could be searching, et cetera, et cetera. And so when we create those skills, then we can package them into use cases that solve problems. So for example, in summarization, we launched a summarization capability based on generative AI that we've integrated into epic discovery, which we talked about.

So that's an e discovery use case, but then it's also usable in the epic service cloud for use cases around deposition summarization and deposition prep.

Launch quite a few services. And you might have seen the announcements we made with both Microsoft around Azure and AWS. But because we, I guess, are cloud native people that happen to come into this industry from prior lives, we've endeavored to be a leader in this industry in the use of cloud based AI APIs and services. So our translation AI based service is up 400% year over year. We've got a transcription service and an entity identification service.

And like everything we do in the epic search cloud, we try to create scaffolding that can be used across multiple use cases. So a big part of what we've done has been to build those solutions and we're just starting to have conversations with customers about renting that private, secure AI platform for folks way beyond Emma 500 that can afford to do it themselves and have the capital and the engineering bench. So we're anticipating, based on very, very preliminary conversations, that some subset of our customers may want to leverage the platform rather than just the services.

Roger Pilse
Yeah, so maybe we could break down those announcements a little bit more. There was, the announcement of, the announcement was providing more AI enabled applications and Dash was built on Microsoft Azure and then also the separate announcement regarding Amazon Web services. So why don't we start with the Azure. Sure, why don't we start whichever one.

Roger Pilse
You want, but yeah, that's a perfect place to start for sure.

Roger Pilse
Okay.

Roger Pilse
Yeah, so Microsoft has been a really close partner of ours for five plus years, and we were named twice as their compliance partner of the year. And so we have a pretty large information governance practice, 70 folks, that's growing 50% a year. And so we've been working with Microsoft. We do a close to 200 projects a year with them around, helping clients implement purview, migrate data from prior storage systems into M 365, implement the various workloads within M 365. So we had a great partnership there. And so with the last six months we extended that. So we've got a AI practice that was featured in that press release around what we call copilot readiness. So we've already sold four of those engagements that we've started and having 20 other conversations. So how do you put the data guardrails in place if you're a corporation in the process of implementing Microsoft Copilot? So private information or information that users shouldn't have access to doesn't show up in a copilot user's results. So that's been a key piece of it. And then because so much of the epic service cloud has been based on Azure, our portal is based on Azure, were a heavy Microsoft Azure power platform, user power bi Azure data factory, etcetera. So it made all the sense in the world to extend that to Azure open AI as well, which is the first LLM we put in the epic service cloud AI platform. And so some of those early services that we've announced, such as text summarization was leveraging Azure OpenAI in our solution.

So that was sort of quick. But those were some of the highlights of the Microsoft announcement.

Roger Pilse
Are you also, I mean, you're helping customers customize how they're going to use copilot and helping them develop that out or.

Roger Pilse
Yeah, so several things. So the most important thing that we're hearing from customers so far is to implement it safely. So that's the first step. As part of that information governance practice of 70 or so people, we've got multiple years of time helping clients implement purview and implement the other aspects of M 365 and put data guardrails around it. So the team's just got a lot of experience. So over the last year it made a lot of sense and we're using it internally actually, as we implement copilot. So it made a lot of sense to work with clients around putting the guardrails around Microsoft Copilot. The other thing you mentioned, which was a great question, is for clients that, yeah, exactly. The customization. So for clients that are implementing or want to implement Microsoft Azure OpenAI, especially law firms, they'll have a great deal of data that they can use to supplement what a horizontal LLM like Microsoft Azure OpenAI can provide. So we've implemented as part of the epic service cloud platform, a retrieval augmented generation capability and a lot of other technologies around it. So for a customer that wants to customize it and leverage all their data and make the prompts more effective and get better answers from just using general LLM horizontal LLM capabilities, we're helping customers customize it and integrate the data that they have and the content they have it to make it more useful to them.

Bob Ambrogi
I'm talking today with Roger Pilse, president and general manager of global legal solutions at Epic, about the company's latest initiatives for building and deploying AI capabilities to enhance the legal services it provides for customers worldwide and to help those customers enhance their own workflows. Theres lots more to get to in the second part of this conversation, but first, please take this opportunity to learn about the sponsors who so generously support this podcast.

Streamline your legal practice with practice Panther Practice Panther is the leader in all in one practice management. Trusted by tens of thousands of lawyers across the nation, practice Panther helps firms simplify their workday by streaming workflows, automating time tracking, accelerating client intake and more, all from one cloud based platform. Find out more about how practice Panther can automate your firm today@practicepanther.com. Dot welcome back to Lawnext, where I'm speaking with Roger Pilse, president and general manager of global legal solutions at EpiC. In this second part of our conversation, we talk more about how EPIC is employing AI to enhance the services it offers clients and help clients integrate AI into their own workflows.

Let's get back to that conversation.

Roger Pilse
Do you customers have a lot of questions about generative AI, in particular about how to use it safely and securely? I mean, are they coming to you with the knowledge that they need to be concerned about this issue, but are they looking for you to you for guidance on how to do that?

Roger Pilse
Yeah, for sure. So the copilot readiness offering has been one of our fastest growing offerings. I would say the quantity of customer interest has been high, but you might be slightly implying it. Maybe there are some subset of customers that were about to implement.

Darn, I didn't know that this was going to be an issue. Thank you. Thank you. Thank you for telling me. So there is a little bit of that. And then within the law firms, what we find. You're probably familiar with the acquisition we did of firemen a couple years ago, which is a great consulting organization, but we were already working with law firms around knowledge management. We were imanage's partner of the year for large law for several years, intranets, extranets, and we had also started a data practice for these law firms. And we're known as sort of sober practical consultants, not hype mongers. So how do you actually take new technology and actually implement it practically? So that has led to multiple engagements with law firms. And so obviously for law firms, Genai is almost an existential thing, frequently is how is it going to affect my business? How do I develop competitive advantage? How do I use our law firms content in combination with a generic LLM to make our folks more productive and to develop a competitive advantage. So the consulting we do often relates to which use cases are better for generative AI. So generative AI has a handful of advantages over prior generations, the conversational search capability, or the ability to summarize or the ability to draft or generate. And so consulting around which, of all the use cases that happen in a law firm, genai can help with, and which less so, like outside counsel guidelines. That's a good one. And then once law firms, we help law firms sort of prioritize the use cases, and then inevitably it's how do you clean up the data for that use case?

So that's a little bit different than the privacy point, but okay, if we care about supplementing search. Oh, my goodness. The data is not set up in a way that will actually help that be very productive. So I'd say the majority of the time, an AI engagement is accompanied by a data management and data strategy engagement. Unfortunately, we already had a data strategy data management practice, so that's been a nice compliment.

Yeah.

Roger Pilse
And do you also hold their hands when they have those existential questions about how this is going to affect my practice?

Roger Pilse
Yeah, we try. We're not presumptuous by any means of, we try to stay in our swim lane and this is what we know and this is what we're good at.

Exactly.

Roger Pilse
Yeah. Are you at this point seeing any kind of game changing applications of this technology in terms of the way it's being deployed, or is it more incremental in terms of being able to provide greater speed or efficiency or cost savings or that sort of thing?

Roger Pilse
Yeah.

Maybe it's more on the latter part of your question, which is probably what you're expecting to hear. So it's certainly in the first inning. We've also worked with consultants on this topic who also work with law firms. And so the prevailing wisdom seems to be that while there are certain use cases drafting as an example, summarization as an example, that the chatbot like examples, the wrapping policies and procedures, it may be in the sort of three to five year timeframe before broader ramp up happens. And I was with a partner at a leading law firm just in the last week. And so that partner and the firm is looking at relativity error and thinking through all the possibilities of generative AI. And then the question comes up. The judge I'm working with is very skeptical about AI.

What about the FTC, or the DOJ, or the opposing counsel? And what case law already exists? What precedent exists? So we're excited about the technical opportunity, but where the ramp up is, it seems like many are saying, in that.

Roger Pilse
Three to five year period, how does working with AWS help advance this? What does that bring to the equation? Why are you working with them?

Roger Pilse
Yeah, so the first thing is the sets of services that are available on top of AWS. So I may have mentioned that epic discovery, the e discovery tool, is based on AWS, and being cloud native, that's how we've gotten tremendous processing speeds, etcetera. The thing, Bob, that's been really exciting is the add on surfaces, right?

The ability for a software company to innovate today based on cloud services. It's ten times what I'd experienced in 15 years in my software experience. So we've launched 22 new offerings in the last five years based on the power of AWS and Azure. It's stunning for us to add on top of core processing and core review, a translation service that, like I mentioned, is up four times.

We did in the first quarter what we did all year last year, or to add the transcription or AWS comprehend for entity identification or cloud based OCR capability, so we don't need to use a vendor. So for us to be able to supplement what we're bringing to customers, I feel like, Bob, it's like 25% the time it would have taken to build all that capability before being able to leverage AWS. AWS is providing a lot of those services for us, and then we're also working with Sagemaker and that will help with a set of prioritizations related to documentation. And then AWS bedrock is also very, very exciting. So they've built a capability to be able to orchestrate multiple large language models. And Mistral might be better for one thing, and anthropic might be better for something else. So the ability to just click in with them and that platform and be able to optimize which LLM we might use for which use case, and then ultimately not just use rag around llms, but fine tune llms with patent data or other type of data, we very much wanted to embrace both partners and then, so in addition to the tech, both parties are providing supplementary engineering resources. So we're super excited recently with AWS that they'll be able to contribute what they call builders. So five builders that will help come alongside our engineers and help them build faster and faster, because those builders are expert at that AWS technology.

And then we also mentioned being in the AWS marketplace. And so one of the things we've loved about the Microsoft relationship is just the seller community has embraced us. And now that we're in the AWS marketplace, which was part of that press release, the AWS community, the sellers are embracing us as well. And if a client is having a hard time retiring, their AWS commit or is almost there, they can retire. The client can retire, their AWS commit with epic purchasing, epic discovery. And so the AWS seller is happy, we're happy, and the client's happy.

Roger Pilse
Yeah, I was poking around on your website and looking at the page for the AI practice group that talks about one of the things that they do is to assist customers in deciding where to integrate AI into their legal and compliance work.

That's more or less a quote off their website, but it made me wonder what are the areas where this is best integrated and where are the areas where maybe AI isn't the right thing to be using?

Roger Pilse
Yeah, I love that. I'm glad you caught that. I love the word integrate because our experience in prior lives, including in e commerce, was AI is best used within a workflow that's getting to an outcome.

And there's a lot of point product tech companies out there, and the key is to be delivering an end to end outcome for clients. A client like a whole workflow or a service. So the fact that of our $600 million business, $500 million of it is essentially a managed service, it helps us click in this AI functionality into an end to end workflow that's providing an outcome. So that summarization capability I mentioned. So summarization is one of the big strengths of generative AI. The fact that we integrated that into an ediscovery tool in a workflow that makes it actually useful, because the paralegal or whoever is using it is actually in there already and can use it as part of a workflow, I think that's super critical. So a large part when we're advising folks is where to use it in workflow. And that AI practice you saw on our website, it's been everything from within a litigation, e discovery, what type of tar to use, what type of Cal to use, where to use, early case assessment, whether it's at the beginning of any discovery or is it related to an investigation. There's a lot of investigations will help with that. Or as I mentioned earlier, we're starting to, Bob, see the shift to the right side of the EDRM with deposition prep and trial prep. So we started to bundle our case insights analytics team with our flexible legal talent, which has been super neat. So we're starting to see what I think is sort of a glimpse of the future, which is essentially AI powered lawyer.

And so now it's actually a lawyer advising the client. And it just so happens that that lawyer is informed by AI. And so, you know, it's the first out of the first inning, I think, with that. But as we, as a, our flexible legal talent business has grown a lot, and our legal staffing and so, you know, it's not clear that AI will replace lawyers, but lawyers who use AI will replace lawyers that don't use AI. And so as we continue to grow that business, which has been one of our faster growing businesses, it's been very kind of provocative to see that tuck in of doing enhanced attorney services and providing lawyers on the right side of the EDRM and deposition prep and trial prep, and then tuck the case insights right behind it. It's been neat to see, I think, a little bit of the glimpse of the future there.

Roger Pilse
Yeah. Is AI potentially a threat to your managed services business?

Roger Pilse
We don't think so. So, like anything, one has to constantly innovate. Right. And so what happened in our managed services business over the last five years is we've had excellent growth. Right. And if you take the subcategory of the business that's managed review, five years ago, we weren't using that much tar or cal or machine learning, and today we use a lot of it. The vast majority of our projects use it, but at the same time, we had 7% plus compound annual growth in that particular area. So. So, Bob, even if a typical project, maybe it's a 1020, 30% less hours, et cetera, but by providing more value to the client, sort of all ships rise. And sometimes I think of it as commerce moving from bricks and mortar to Amazon. But we've seen as the more we integrate AI previously ML and next generative AI, it has a corresponding effect of more volume coming our way and the business. So I think on a per project basis, you might see fewer hours, but we're also increasingly pricing per outcome. And so when we price per outcome, number of docs, or a total matter, then we're the recipients of the benefits of the efficiency. The customer gets certainty. But the more AI we use, the more it drives our profits. And our profits have been growing as well.

Roger Pilse
That makes sense. Just curious. I mean, I've seen a lot of debate online about the extent to which generative AI is how it performs relative to tar or machine learning in document review, and which is better. Do you have any thoughts on that or how they compare to each other?

Roger Pilse
Yeah, it's a good question. It's definitely early inning. So, that is one conversation we're having. I mentioned a partner at a large law firm, and we brought one of our best consultants in, and she's got a lot of experience in Tara, Cal, and now she knows generative AI and relair a great deal. And so that's. I would say she's one of the smartest people in the industry. And it was sort of. It was an inconclusive conversation, I would say. So there's. I'd say there is some conversation of promise. The thing, Bob, that jumped out in that conversation, like many, is not just the will the judge accept it? Will the courts accept it? Will the regulator accept it? It's the cost. And so the main feedback that that part partner provided is, I see some promise here, it could be very productive, but let's keep an eye on it because at this point the costs are relatively high. So I think service providers like ourselves, a key value that we'll be providing for clients is helping clients optimize costs, and it might be using machine learning for a large part of it because that's much cheaper, but then using generative AI for a subset which will be more expensive. And I think often clients will receive a composition of services that ideally gets them the best outcome at the lowest cost. I think a big variable right now is just the sheer cost of using the Azure, OpenAI and other platforms. And so all of us, whether we're tech companies or service companies, are now kind of intermediaries between the customers who want value for a good cost and a high cost basis.

So I think that'll be the other part over the years that'll be necessary for those curves to start to change before more widespread adoption can happen.

Roger Pilse
Right, and that's probably where it makes sense to be figuring out which models are best for which use cases and precisely use the one you can use the most affordable when you can.

Roger Pilse
Exactly, Bob, you nailed it. So that's why we ended up building an AI platform that was LLM agnostic, and these LLMs could also even be in our physical data centers as well. So as we've started to work through the various use cases, we're already seeing opportunities to optimize for our customers so that they get the most, you know, most legal intelligence, as we call it, for the, for the lease cost.

Roger Pilse
Roger, can you give us any other glimpses of what's on the radar for you for epic over the next year or so?

Roger Pilse
Yeah, for sure, Bob. So within the generative AI based solution. So there'll be more coming later in the year than the ones we've already announced. So there'll be more in the area of AI based classification, there'll be more in the area of AI based search. I mentioned it briefly, but a conversational digital assistant at Chatbot, as people say, so that will be in customers hands towards the end of the year. So we'll see more and more of those AI based services coming out. We're onto version two of our AI platform, and so begin to have conversations with customers about that.

Gen three will be available in the third quarter, and then from a consulting standpoint, I think what we're offering today will carry us for a lot of the year. So I think helping the law firms with their GNAI strategy and their data strategy and corporates with their co pilot strategy, I think we've got our kind of hands full with what we're offering today. But one thing you and I touched on quickly about AI powered lawyers. So I do think that that example of flexible legal talent plus case insights, I think as the year progresses and our flexible legal talent business grows, that might not be a 2024 second half thing, but I think in the first half of 2025, as that organization continues to grow, having more and more AI empowerment of that flexible legal talent, like ultimately we're a billion two alternative legal service provider. So I think more and more of that flexible legal talent business and our lawyer secondi business and the use cases we're doing, whether it's the deposition prep or other things, will be powered by AI. So we're really excited about that.

Roger Pilse
Roger, thanks so much for speaking with me today. It's been a pleasure to have you.

Roger Pilse
It's been a pleasure being here. Bob, thanks so much for having me.

Bob Ambrogi
My sincere thanks to Roger Pilse, president and general manager of global legal solutions at Epic, for joining me today. I hope you enjoyed the conversation.

If you'd like to share your own thoughts or comments on today's show, please do so by messaging me on LinkedIn or X or whatever social media platform you're on or by emailing me directly@ambrogiomail.com.

And if you're a fan of Longx, please leave us a review wherever you get your podcasts. Longx is a production of Longx Media. I'm your host, Bob Ambrogi. I hope you'll join us again next week. Time for another episode of Law Next.