Insights

Read the transcript below:

The following episode of Beyond Hourly was recorded before the merger of IMF Bentham and Omni Bridgeway.

Allison Chock:
Hello and thank you for tuning into the Beyond Hourly podcast hosted by Bentham IMF (now known as Omni Bridgeway). We focus on advancements in legal services that drive economic value for law firms and the clients they serve. As one of the world's most experienced commercial litigation funders, Bentham IMF has invested nearly two decades into providing litigation finance and investment capital to plaintiffs and law firms for large disputes. We offer law firms and their clients a risk-sharing partnership and a proven record of success as a leading global litigation funder. Episodes of this podcast can be found on our website, www.benthamimf.com (now at www.omnibridgeway.com), iTunes, and other podcast networks. We welcome you to subscribe to the podcast and leave us reviews.  

I'm your host today, Allison Chock. I'm the Chief Investment Officer of the U.S. Operations of Bentham IMF. Before joining Bentham IMF in 2013, I was a practicing lawyer for 14 years and I did a lot of contingency trial work in the security space. My role at Bentham now involves overseeing Bentham IMF's investment management and due diligence processes across its four U.S. offices.

My guest today is Enoch Liang, who is the Co-Founder of LegalMation. He also was the Co-Founder of the law firm LTL Attorneys LLP. After litigating at Quinn Emanuel and then practicing in China for three years, in 2008 Enoch founded LTL, which he helped to grow to nearly 40 lawyers over the next decade. In 2018, Enoch co-founded LegalMation, which is a company that leverages artificial intelligence to automate litigation process tasks. Enoch is currently the Chief Strategy Officer at LegalMation.

Welcome, Enoch.

Enoch Liang: 
Hi, Allison.

Allison Chock: 
Hi. Thank you again for coming on, it's really great. You and I have known each other for a few years. We  
won't go into all the ways that we've known each other through the years. It's great to have you on and I congratulate you on the great success of this new product.

Enoch Liang: 
Thank you. We won't talk about the late night Halloweens in the past as associates.

Allison Chock: 
No, we don’t want to talk about that. If you would please give us a brief description of the LegalMation tool and its capabilities.

Enoch Liang: 
LegalMation, as you mentioned, uses artificial intelligence to automate those processes involving litigation tasks that all of us former litigators know about and hated doing when we were associates. For example, our first product—which came out into production I think in April of 2018—all you do is you take a PDF of a complaint, you drag and drop it into our engine and our engine will OCR it and then send it to IBM Watson. IBM Watson will analyze it and then in two minutes you'll have an initial draft answer, affirmative defenses, as well as targeted requests for production and interrogatories, and if you also want a deposition notice, It can do that in two minutes. I think when I was young associate it probably would have taken me eight to 12 hours to do that initial bundle work.

Allison Chock: 
That's pretty quick work actually. It would have taken me longer. For those who haven't seen this tool in action, it's quite phenomenal. I saw it and was completely blown away myself. I said immediately, what the heck did I do with the first six years of my life practicing law? Because it just wiped that out. I think it's really quite fabulous. What else does it do now? Are there other features?

Enoch Liang: 
Yes. That was the first product that came into production. You know, that first set up of work that I just described happens once in a case. We wanted to help litigators throughout the timeline of a case. The next thing I think that happens in litigation is after you've served discovery, you have to be able to respond to discovery. One thing I hated doing as an associate was, the other side serves discovery on you, and before you can even get to the substantive work, you actually have to spend time coming up with the shell responses and the objections to each and every request. If you get served 240 requests for production, that could take literally a day or more just coming up with the objections and we thought there's no reason why we can't train artificial intelligence to do that.

Our next product, which I think launched in September of 2018 is basically when the opposing party serves discovery requests on you, you do the same thing. You drag and drop it into our engine—and this one's faster. I'd say within a minute it'll spit out for you the responsive shell with all the general objections at the top. It'll have each request. It'll have the response to each request with the artificial intelligence suggested objections to each and every request. Of course, we allow lawyers to customize as well. If there’s something custom about that particular set of discovery that you should object to, such as it violates a protective order, you can go ahead and type that in and the engine would automatically put that into the response for each and every request. I'd say that that takes four to six hours of work and turns it into one minute.

Allison Chock: 
It is really, really impressive. What was the light bulb moment that made you decide that you needed to create this LegalMation product?

Enoch Liang:   
I think as you mentioned in the introduction, we had our own litigation boutique for nearly a decade. I think the light bulb moment was James Lee—who's also a Co-Founder of LTL and also a Co-Founder of LegalMation—he attended the Harvard Law School Executive Education Program for Law Firm Managers. It's a week-long program. It's great, and if you've never heard about it, I highly recommend it, especially if you're in law firm management. It's from Sunday until Saturday, 10 hours a day of classes followed by networking events at night. I think it was Thursday afternoon, there was an entire module of technology. I don't think it's any surprise to say that lawyers are generally the last to adopt technology.

Professional services firms, financial services firms, even doctors have been starting to use artificial intelligence in their work. That whole module at Harvard for three hours was talking about how technology is changing those fields.  James came back from that week long class and basically told the firm, "Look, we need to explore how we can use artificial intelligence in our own practice, because if we don't, the AmLaw 200 is going to explore it and end up crushing us." We were already a small firm at the time. We were 40 attorneys. We were constantly litigating against armies of lawyers from like Jones Day, Hogan Lovells, O'Melveny, MoFo, etc. We decided to pour some resources into exploring whether artificial intelligence could be used to make our lawyers more effective. That was kind of the light bulb moment.

Allison Chock:   
Yes, right. We've had similar revelations here about using artificial intelligence. We use it sometimes to help us due diligence cases. It doesn't replace what we think is our professional judgment, but certainly is a great assistive tool.

Enoch Liang: 
It's absolutely a great supplement. In the end I think you know, AI isn't going to be writing Supreme Court briefs or standing up before a jury or a judge, but getting all that routine work out of the way that none of us went to law school to do is I think quite valuable.

Allison Chock:
Next, I'd like to turn to sort of, from the light bulb moment how you guys made it happen. Are you a coder, or is one of you a coder and how did you get started?

Enoch Liang: 
LTL itself had often focused on software litigation, software licensing disputes, etc. As a result, we had hired, and we'd made a focus of hiring, associates that had a computer science background. There were several, I think there were two associates that were either self-taught programmers or had a computer science background already at LTL. We formed a small group to kind of explore the AI that was available out there and explore applications that AI might be suitable for because AI isn't suitable for everything.

I think we did that for about six months, and we were getting nowhere until we decided to kind of start doing more supervised training. That's when things actually started to work out and we realized that AI could actually do some of these process volume tasks. Once we did that and we showed some of LTLs clients, that's when LTLs clients said, "Hey, you guys are great and all, but you're only 40 attorneys in California. This tool is so valuable. We'd like to use that nationwide." We ended up spinning off LegalMation out of the law firm. I think we incorporated January of 2018, maybe December 2017 and now have the separate standalone business that's just licensing software.

Allison Chock:   
Wow. How many do you now have on staff at LegalMation?

Enoch Liang: 
I think we have something like 27 to 30 people. Of those 10 are coders and 20 are attorneys, because we found that it takes something like that two-to-one ratio in order to come up with products that attorneys would actually use.

Allison Chock: 
That makes sense. What has been the biggest programming challenge that you've encountered to date?

Enoch Liang: 
I would have to say the artificial intelligence supervised training. It is what it is because generally speaking, attorneys in California, attorneys in Florida, attorneys in New York are going to draft legal documents relatively the same way, right? In terms of the language. But the formatting and the templates are totally different. Between let's say Central District of California and Southern District of New York, between Superior Court Los Angeles County and a similar state level court in Florida. I would say that the biggest challenge for us has been in the OCR parsing aspect from state to state, jurisdiction to jurisdiction, as well as the document generation templates from state to state, jurisdiction to jurisdiction. While the AI training is the same for the actual substance, to get documents that look like some something an attorney would use in each one of those states takes significant amounts of programming and training.

Allison Chock: 
Then how has the marketing of LegalMation, of the software tool, proven different from—obviously you are great marketer of the law firm business—do you sort of do it together or are they two separate enterprises from the marketing perspective? How does that work?

Enoch Liang: 
I mean definitely LTL and LegalMation are two separate enterprises. Two private entities. In terms of the marketing, I would think that LegalMation's marketing is quite different than marketing for legal services, because legal services marketing is I think pretty well established in terms of how you're going to market a law firm, how you're going to get business for a law firm. I think legal technology marketing is a kind of new space that's only arisen in the last five years or so. Convincing law firms, enterprises, insurance companies to take on new legal technology is a very different proposition than trying to get a case.

Allison Chock: 
Who has been buying into the product so far?

Enoch Liang: 
There's a lot of law firms and companies that are under NDA, but the public ones that we can announce are Walmart and Ogletree Deakins. We also have a number of insurance companies that, like I said, are under NDA. If you were to think about the top five insurance companies that come to mind, chances are we're probably working with at least three of them.

Allison Chock:   
Right. I did a panel on technology in the law actually and generally on disruption in the law, and I was describing your product without naming it and someone else on the panel leaned over and said, "Is that LegalMation?" He's like, "It's amazing!" It's definitely getting out there. We've talked about how this is essentially turning what we used to do as junior associates and hours and hours of work into something that can be done in seconds. What's been the reaction of associates who have seen the tool? Are they worried about it taking away from their billable hours, or is it more elation at the prospect of not having to spend hours and days on kind of rote work?

Enoch Liang: 
I think it's both, and I think that the reaction depends on the law firm's model. Your podcast is called Beyond Hourly, which I think is the way that the legal market is moving towards. If you have a firm who 99% of their work or 90% of the work is on an hourly basis, generally speaking, they're not going to be that receptive to LegalMation. Because the incentives are misaligned. The firm is not incentivized for being efficient. Why take something that can be done in 12 hours and do it in 30 seconds, when you could bill for the 12 hours. However, if the firm does a significant amount of flat rate work, either on the plaintiff's side or the defense side, or contingency work on the plaintiff's side, they are very receptive to LegalMation, because if you're doing work on a flat rate basis or contingency basis, then you have every reason to be as efficient as possible. The more inefficient you are, the less profitable you are. I think the reception really depends on kind of what the firm’s business model is.

Allison Chock: 
Does that carry through whether it's associates or partners? Is it the same, it depends upon what the firms leverage model looks like?

Enoch Liang: 
Yes. I will say that the complaint analysis tool, while people like it, especially associates, they are much more excited about the discovery analysis tool, because discovery analysis is something that's really, really painful for associates and paralegals to deal with.

I'll give you an example. One of the firms we were working with, they had a case with 20 defendants and the 20 defendants each served 60 sets of RFPs on the plaintiff. That's 1,200 requests. In the past, if you were to receive 1,200 discovery requests, the playbook would be let's go to the court, get a protective order and have the court limit the discovery. That in and of itself takes a lot of time. Today with LegalMation, we were able to take the 1,200 discovery requests and convert them into the shells and have a draft stock response and a draft stock of custom objection in seven minutes to 1,200 requests. I think that it's changing the way … LegalMation is going to change the way that people are practicing. You can imagine as an associate or a paralegal how much you would love to have access to a tool like that.

Allison Chock:   
I know that some large companies have actually bought the LegalMation tool to assist their legal departments with bringing some of the tasks which are traditionally performed by junior lawyers into their in-house legal departments. You've mentioned Walmart for one. How's that going so far? Has the corporate client's experience with the tool been different than law firms?

Enoch Liang: 
I think it's been going quite well. We recently completed an analysis for a large corporate customer, and it showed that the LegalMation tool was saving 80% of the time. 70% to 80% of the time that it would otherwise be taking to complete the tasks. I think that there's more and more of a trend, large corporates taking certain legal work in-house. I think that you're going to see a trend of kind of disaggregation of legal services, deep bundling, so to speak. For example, we have one client that they run the LegalMation work product in-house using their own junior attorneys and paralegals, and they send it to their outside counsel together with the complaint and basically say, "You can only bill us now for reviewing or revising this work product, because we've given you a good first draft." I think you're going to see more and more of that as you see the rise of alternative legal service providers, legal tech companies, changes like that to the legal profession that are going to de-bundle or disaggregate legal services.

Allison Chock: 
What has been your favorite reaction so far from a client?

Enoch Liang:   
I think that we were giving a demo to a law firm and what we had done was we had actually taken a complaint that had been filed, let's say two weeks ago and we knew this law firm was the firm assigned to defend it, and we ran the LegalMation output on the spot during the demo and went through it together with all the partners and associates that were sitting there at the lunchtime demo. One of the people in the back started clapping because she, it was a partner and she said, "That is my case and I can't believe that you guys just did that in two minutes. That's amazing." She just started clapping and as you can imagine, we signed a contract that day.

Allison Chock: 
That's fantastic. What do you see in the future for LegalMation? Do you think its geographic expansion? Are you continuing to develop different tools within the product?

Enoch Liang: 
It's both. Geographically we are now in California, Texas, Florida, New Jersey, New York, and Illinois. We're basically following state-by-state in terms of the litigation volume. We're going to continue, we'll probably build out the top 20 states, because of the top 20 states will probably have something like 90% of the litigation that's filed in the United States. Geographically, we're going to expand. We're going to expand into different areas of law. Currently, we have an employment law domain, a personal injury law domain, an insurance law domain, an asbestos domain, a medical malpractice domain. I think we want to expand it to commercial litigation, maybe soft IP, those types of domains. We have to expand geographically; we have to expand domains of law.

We also want to expand along the timeline of litigation. Like I said, most litigators know this. Every case has basically the same lifecycle. There's the complaint filed, then the answer filed, then the parties exchange discovery, and then summary judgment, then pretrial, then trial, and then appeal. We think that we can build tools to make lawyers lives easier along that entire timeline of the case. For example, right now we're working on a third module, which is similar to the discovery analyzer, but right now the discovery analyzer gives you the AI suggested objections along with the custom objections that you want to put in. There is no reason why for companies or law firms that are facing the same types of cases over and over again and are facing  similar questions, there's no reason why the AI can't sort and come up with client approved answers and do the first draft of the discovery responses themselves. We're working on that.

We're also working on a data analytics platform. We're already in a pilot on a data analytics platform with several large companies and several large law firms. The idea is that currently most data analytics platforms suffer from a couple of limitations, right? The first limitation is that if you're building an outcome analyzer without access to settlements, then you're missing 98% of outcomes, right? Only 2% of the cases are going to trials. That's one limitation. The other limitation is that many data analytics platforms are very, very high level. If you look at the pacer nature of suit codes, those are super high level, right? They don't tell you that much about the facts of the case. Because of what we're able to do with our complaint analyzer, we are actually able to tell you the exact allegations that each complaint is making.

It's not just for example, age discrimination. It would be, we're able to tell you that it's an age discrimination case where the plaintiff is 55 plus, which is going to be different than a case where the plaintiff is only 45, right? Where the supervisor was 32 and where the plaintiff had worked there for 15 years. We're able to give you the really, really deep dive that we feel like is the master key to unlocking the data insights down the road. Everybody knows what the billing data looks like, right? Companies have that, law firms have that. Most people will have the sum of outcomes. Companies have that, law firms have that. What they don't have is they don't have the front end analyzer to figure out how to thinly slice that data into actionable intelligence. We're pretty excited about that.

Allison Chock:   
My request is for you to finish that tool faster and we will help you test it out.

Enoch Liang:   
I could imagine why Bentham would like that.

Allison Chock:   
Absolutely. Do you think the LegalMation tool could ... I mean it sounds like you've got a lot in your future vision already, but could the tool be tweaked to be expanded outside the legal industry? I was thinking sort of maybe in like real estate transactions or something like that. Or are you just trying to keep your eye on the ball?

Enoch Liang:   
The answer is yes it can. In fact, we've had clients ask us to take our data analytics platform into the regulatory and compliance already. We've also had clients ask us to move into the transactional space. From our perspective, there's a couple of reasons why we've been focused on litigation, and hyper-focused on litigation. The first is we are not subject matter experts in transactional or regulatory or compliance. We are litigators and so we understand what litigators need and what litigators would use. If we were to expand into those other fields, I think it would require much more subject matter expertise than we have right now.

The second reason is that there aren't that many competitors that are focused on the litigation space, especially in terms of automating litigation processes and litigation drafting. There are a ton of companies trying to use AI to solve pain points in the transactional space and in the regulatory space and in the compliance space. I think we have a niche right now and we have competitive advantage in terms of being first to market in the litigation space and we wouldn't want to lose focus and focus on something else that we're not subject matter experts in, where there's more competition.

Allison Chock: 
The last thing you want to do is get distracted and launch a magazine or something out of the box. Well, are there any other AI assisted tools that you actually admire or think are also worthwhile?

Enoch Liang:   
I would say that AI assisted legal research has come a long way. You know on the LTL side, I know that we've tested Casetext and their brief analyzer product and you know, we've been impressed by that. We've also seen that Westlaw and Lexis and others have come out with similar products in the last two months. I thought it was just a great idea to be doing that because that's exactly what lawyers need, right? You upload your brief and it tells you all the best pro and con cases that you didn't cite. That's exactly, I think, what young associates need and what partners need to know when they're reviewing a brief. Is there anything, are there any lurking mines out there that we haven't uncovered, right? I've been impressed with that tool.

I think I've been impressed with some of the litigation analytics tools out there that are already being offered by, let's say Westlaw or Lexis or other ones like Catalytics. Again, like I said, I think one of the drawbacks of those tools is you're not getting 98% of the results. You may be able to get things like mean time to resolution for this particular judge or this judge rules summary judgment in favor of defendants 75% of the time. Okay, that is helpful information, but I think that most of the actionable intelligence for companies is going to be at the outcome and billing data results that only they have behind their firewall. I think that I've been impressed with those tools, but somebody needs to convince clients and law firms to share maybe on an anonymized basis that data that only they have in order to build a tool that's truly going to lead to insights and actionable intelligence for the business side as opposed to just litigators.

Allison Chock:
That's definitely been our experience as well. It's the data input, making some of these predictive tools, maybe not so predictive, because if it's garbage in, it's going to be garbage out. If you're missing a whole lot of data, then you're not going to get terribly good predictions out of a tool, no matter how great the AI is.

Enoch Liang:   
I think that's what really excites us about using the LegalMation AI tools for data analytics. Right now, how are people inputting the matter profile? I mean, basically you're having a secretary or a paralegal or an associate do it. They're not able to bill that time. Right? They're not updating it later on once the case is resolved, because they're too busy moving onto the next case. In a sense, our competition at the matter profiling for data analytics space is really other humans, whereas we're able to, at scale, automatically generate really detailed matter profiles. In other words, the digital fingerprint for each and every case that's uploaded into our system. That really is the master key that will unlock the actionable insights from the billing data and the settlement data.

Allison Chock: 
You're also a named partner at a successful boutique litigation firm. How do you manage to do all this?

Enoch Liang:   
I would have to say that we've taken a significant step back from the daily practice of law, as well as the law firm, and have essentially turned it over to a younger generation of partners who are more focused on maintaining the excellence over at LTL. They've been doing a great job. We've been more focused in terms of our day-to-day on LegalMation. I would say that in terms of daily litigation practice, I have one case left. I consult and give high level strategic advice on a number of cases. In terms of the actual day-to-day stuff, I have one case left. I think the same is true for James, and so we're pretty hyper-focused on building LegalMation.

Allison Chock:
Suffice it to say, if LegalMation is the next unicorn, you will happily give up the practice of law? You're a great lawyer, I think that would be a loss for the legal industry, but I suppose some could argue there's plenty of lawyers out there already.

Enoch Liang:   
I don't know if I would give up the practice of law altogether because I do still think that one thing I miss about being so hyper-focused on LegalMation is, I miss crushing the other side. I kind of miss that thrill of victory. There's a flip side to that coin, which is the agony of defeat, right? All the work that goes into litigating a case all the way through trial and then all the way through appeal. I would think I'd always have my fingers involved somehow in litigation because I enjoy the puzzle solving aspect and I enjoy the competitive aspect of it. Hopefully if LegalMation is a success, which I think it will be, we'll make the practice of litigation much more palatable and free up litigators to really focus on what matters, as opposed to all that processed volume stuff that you're stuck doing your first three or four years at a law firm.

Allison Chock:   
Well thank you everyone for sticking with us. I'm going to give you the website address for LegalMation, it is www.legalmation.com and you can find Enoch on LinkedIn under Enoch Liang. E-N-O-C-H L-I-A-N-G. Enoch, I want to thank you again for appearing on Bentham's Beyond Hourly podcast and sharing your knowledge.

Enoch Liang: 
Thank you, Allison.

Allison Chock:
Everyone, as I mentioned at the outset, episodes of the Beyond Hourly podcast can be found on our website, www.bethamimf.com (now at www.omnibridgeway.com), on iTunes and other podcast networks. We'll be back soon with another episode focused on advancements in legal services that drive economic value for law firms and the clients they serve. Until then, I'd like to thank you all for listening in and invite you to subscribe to the podcast and leave us reviews.

Feel free to follow up with me anytime, Allison Chock at [email protected] (now at [email protected]) for any feedback, ideas, or insights you have on topics that we cover on the podcast. Thank you and be well.