In this episode of The Geek in Review podcast, hosts Marlene Gebauer and Greg Lambert welcome Jeroen Thierens, Strategic Account Advisor, and Jorn Vanysacker, co-founder of Henchman, a Belgian legal tech company that focuses on building an intelligent drafting assistant for lawyers working on complex transactional contracts based on precedents in the firm’s DMS.
Vanysacker shares the story behind Henchman’s founding, emphasizing the importance of solving a clear problem with a focused product strategy. The company’s mission is to unlock the collective knowledge within law firms and legal departments by connecting to their document management systems, recognizing contracts and clauses, and making this information easily accessible within Microsoft Word and Outlook.
Henchman serves both in-house legal teams and law firms, with a customer base spanning from boutique firms to AmLaw 200 firms. The platform addresses two main use cases: drafting transactions and assisting knowledge management teams in gaining data-driven insights to prioritize their work and boost relevant knowledge in search results.
The discussion also touches on Henchman’s integration with Microsoft Copilot, which aims to provide more accurate results when drafting contracts by leveraging the structured data and metadata within the platform. The company’s European perspective is reflected in its language-agnostic system, which works seamlessly with multiple languages, making it an attractive solution for global law firms.
Looking ahead, Vanysacker and Thierens discuss the challenges and opportunities presented by generative AI in the legal industry. While acknowledging the potential of large language models, they emphasize the importance of context and accuracy in transactional work. Henchman’s approach is to use AI as a tool to enhance their product, focusing on providing the next best options rather than just the next best words, ultimately aiming to become the “Chief of Staff” for each legal professional.
Listen on mobile platforms: Apple Podcasts | Spotify | YouTube
Contact Us:
Twitter: @gebauerm, or @glambert
Threads: @glambertpod or @gebauerm66
Email: geekinreviewpodcast@gmail.com
Music: Jerry David DeCicca
Transcript
Marlene Gebauer 0:00
This podcast focused, innovative and creative ideas. I’m more than
Jeroen Thierens 0:14
And I’m Greg Lambert.
Marlene Gebauer 0:16
As you could tell from the opening music, this is one of our regular shows where we look into the industry and find innovative people to bring on and discuss what they’re doing in in advancing the legal profession. We’re
Jeroen Thierens 0:27
really excited to welcome Jeroen Thierens, Strategic Account Advisor and Jorn Vanysacker, co founder from Henchman Jeroen and Jorn, welcome to The Geek in Review.
Yeah, thank you for having us.
Marlene Gebauer 0:43
So you’re in before we jump into some of the really great things that you’re doing and the growth of Henchman in the past couple of years. Could we get you to give us a brief overview of Henchman and why you found it?
Jorn Vanysacker 0:54
Yeah, of course, Marlene. So hey, I’m Jorn . I’m one of the three co founders at Henchman today. We’re 40 Very talented, techies together building Henchman. We started in 2020. And yeah, we’re based out of Belgium, but we have more than 170 customers in more than 30 countries of which the US is the biggest country that we serve at the moment. And it all started, I mean, Gilles [Mattelin], myself, we had a previous company, which was building HR tech, and back in 2013 2014. And yeah, I mean, we were building a suite solution, we were building something for performance management reviews for employee engagement measurement at the time. And I remember we had one competitor called Peakon. And they just did employee engagement in a very data driven manner, very, very kind of vertical in terms of product strategy, we were really inspired by them. And they ended up selling to Workday for $800 million, something like that. And we got acquired in 2019, for a very small fraction of that amount at the time. But Gilles and myself, we had a very clear idea of like, okay, if we want to do something next, what’s our next project going to be? And it had to be something that solved a very clear problem, and must have products had to say, and with a very strong focus in terms of what we wanted to solve and for who we wanted to solve it. So we went up into the direction of okay, well, as you had to fill in a lot of RFPs. And he was very frustrated with the fact that you always get these long lists of questions. And he remembered that he has given very similar answers, like let’s say, two months ago. And so he thought of an idea of like, let’s create a technology that could recognize similar answers I’ve given to similar questions from my database. And so we had lunch with the legal professionals that assisted us in the acquisition of the previous company. And they were like, Guys, this is what we do on a daily basis, like looking for stuff we’ve done in the past. And so in that lunch meeting, we pivoted to the entire idea of Henchman to legal and that’s how we got into legal. And so what Henchman does is built for legal professionals today that draft and negotiate complex contracts, the transactional kinds. And the two challenges we solve for them are the first of all, when drafting these complex contracts, they tend to open up a lot of contracts, a lot of folders looking for that one clause they have in mind, and you know, if they don’t find that they ask a colleague, have you ever done something like this? And then the second thing, and it’s more of an opportunity, really, is the fact that we, you know, they all these legal firms are these organizations that sit on a lot of contracts a lot of data, and it’s often very unstructured. And so with the right tech, you can really use that data as a lever for your, you know, legal colleagues, but also, of course, your clients. And that’s really what we’re obsessing about with Henchman unlocking that collective knowledge.
Jeroen Thierens 4:02
I have to have to ask, who’s the harder industry to deal with HR or legal?
Jorn Vanysacker 4:08
I was, I was I, it’s a good question. I was a bit of fried in the beginning to be honest of, you know, legal professionals and you know, would they be mean, and but I have to say, I like them a lot more than then HR professionals, because they’re men, especially the transactional kind of lawyers. That’s the ones that we work with the most today. They’re very constructive in nature, right there deal makers. And in that sense, it’s been a real joy building Henchman together with them. Yeah.
Jeroen Thierens 4:39
I would much rather talk to a to a transactional lawyer than to the HR team. So sorry, you were gonna go on?
Jorn Vanysacker 4:50
Yeah. Well, I mean, I think when when we started off Henchman, we were product people, right? We’re tech people. We were not legal people for as founders, so we were forced to really understand the problem very well. And so we still do a lot of workshops on a monthly basis with our customers and prospects. But two things became very clear to us during those conversations. So first of all, legal professionals are very, very busy people. And second, and I say this with the utmost respect, they’re not too tech savvy. And so whatever you’re building in legal right has to be something that works for them, rather than something that has to be set up by them in order to get value out of it. And so that’s really been our main objective when building Henchman and so the way we built Henchman works for our customers is that first of all we connect to their document management systems can be iManage, NetDocuments, SharePoint does matter. And we will just plug and play, there is no setup required no preparation of folder structuring requirements. And then the second phase Henchman recognizes what are contracts and what are not. And then Henchman will isolate all the call clauses and definitions from those contracts, it will mirror the permissions that are in place in the DMS, we will recognize the associated metadata to those contracts. And we will create metadata on the process and all of those clauses like per customer, we easily have a couple of million of clauses and definitions that we process. All of those millions of clauses and definitions are accessible within Microsoft Word or Outlook. And so yeah, if you look at our customer base, I would say 20% of our revenue is in house legal teams and 80% is law firms with a very widespread from, let’s say, boutique firms to Yeah, you know, AmLaw 200 firms.
Marlene Gebauer 6:41
So I’m curious, you know, in working with firms, are you hearing, you know, what types of reasons, are they sort of needing the clauses surfaced? I mean, is it primarily like we need precedent to, because based on what we sort of have done before, are they looking? Or they’re sort of more, is there more sort of an analytic bent to it like saying, Okay, what have we done in certain instances? And, you know, what have we agreed to in certain cases where we maybe haven’t and others? And why? Anything like that?
Jeroen Thierens 7:16
Maybe I can think that question, Jorn, talking to a lot of law firms. I think it’s indeed, definitely also the letter, Marlene, there’s a lot of data in there. Indeed, that is now not really accessible or don’t have a lot of insights in that. And indeed, Henchman surfaces that knowledge and with all the metadata around that, it’s very easy to get a better understanding of how we did similar things in the past that we can apply now. For example, in transactions, it’s always compared a bit with a chess game. It’s always nice if you know if it’s what you can do a few steps ahead. It’s the same. And if you indeed can leverage Henchman and have insights in how we’ve done similar things in the past, then can really help you in this current transaction as well. So we see this really resonates a lot with our customers and our and our prospects. So that’s one thing. Actually, I see two main use cases. And while there’s one very major use case, this is all about the drafting the transactions, what I just explained. But also we see a lot of interest from people and dealing with knowledge management, in Henchman. Yeah, of course, we are originally a Belgian company, we focused on Europe first before crossing the pond to the US. And yeah, of course, in the in Europe, Knowledge management is very established for a long time already, especially in UK firms. But also in the Nordics. I focused a lot on the Nordics in the past. And as Jorn also said earlier, we focus heavily on doing workshops with our customers, just this week, our product team went to Stockholm in Sweden to have a product workshop just on knowledge management. And then this gives some very nice insights and how we can further build our product as well. So also for knowledge managers, or PSLs Henchman is aiming to give a lot of insights, data that helps them in their job, for example, it can help them when they see that certain clause are searched for a lot, they can prioritize their work through that kind of clauses. Whereas if certain contracts are rarely searched for then you know, as a knowledge management team, okay, this is not the type of work that our customers do a lot. So our customers, or our lawyers do a lot. So let’s not put too much time on that. It really helps also in strategy, strategizing, where to put your time and efforts as a team and yeah, to put it a bit more broadly. Yeah, when talking to US law firms now at least, a big part of them indeed has a knowledge management team in place, but that’s definitely not the case for Are all of them. Some of them haven’t always focused on knowledge management. But I see this is really changing. A lot of the AmLaw 200 firms we’re talking to attention, really want to step up their game when it comes to knowledge management. And they’re looking at two things in that regard, either hiring more KM people, but definitely also looking at technology to further automate this. And they’re often looking at a tool like Henchman to do this, because, indeed, we give them a lot of insights, data driven insights in their current knowledge. Yeah, to to really stimulate that knowledge work. And with the insights they gather from Henchman, they can also make sure that certain knowledge get boosted in the search of Henchman. That’s lawyers see this first. So yeah, we really see that there’s a lot of interests also from a KM perspective in Henchman when it comes to that.
Yeah, let me let me pick up on the the technology because I and we’ll get back to the the Europe versus us a little later. But a lot of US KM teams look at km as a technology, whereas it’s more of a personnel kind of a process in Europe. But I’m gonna want to focus in on the the US centric and that’s the leveraging the technology to to help with the knowledge management. And I know you guys are using Microsoft Copilot to help with the drafting experience. So for the legal professionals that are using the product, and specifically, you are looking at the firm’s document management system, or the corporation’s document management system, if they’re lucky enough to have one. You know, and so for us km folks, this is kind of the holy grail that we know there’s a lot of data inside our DMS that we would like to leverage to improve our km process. So how are you guys using Copilot to help with that?
Yeah, sure. So indeed, I would situate copilot more in the first use case I explained to the where legal professionals, lawyers, practicing lawyers are using co-pilot. The reason why we start building that integration we currently have with copilot is actually because yeah, in those conversations I, for example, have with AmLaw 100 firms, the big majority of them, if I may say so is experimenting with with Co-pilots. And what I hear there quite often is that it’s a great productivity tool. It really helps in both Outlook, drafting emails or in Word, Excel is great for general productivity. And that’s, of course, always the goal of Microsoft, they always want to stimulate general productivity. But when it comes to the more, deep legal work, if I may say so such as drafting contracts, it’s not, it’s not there yet. With the co-pilot, what I hear quite often is indeed, when you give a prompt, for example, give me 10 liability clauses written in an SBA by this specific colleague in the last year, then co-pilots will give you some kind of answer based on your data. But it’s not always accurate. For example, if you then open these sources, those contracts, sometimes there is not even a correct liability clause, and they’re the one you want it. So it’s not working that well when it comes to that and what Henchamn is really good at and what you’re already briefly touched upon earlier. We’re good at in that indexing those clause definitions coming from the document management system, we structure them in a very good way in our in our software, together with all the relevant metadata attached to that. And that really helps co pilots because the integration as we build it right now is that when people make a promise in CO-pilot in the like, give me them liability clauses written in an SBA by Marlene, for example, in the last year, then cooperate with make an API call to Henchman and Henchman, which returned results to the user through co pilot. And then of course, you have way more accurate results than when it’s just co-pilot that you’re using, again, because we have so much metadata attached to those clauses that are structured very well. So yeah, we see that this resonates very well in the last weeks. And we also hosted a webinar on this together with our partners, iManage and Microsoft, and yeah, your that was crazy, right? We did it in the Microsoft HQ here in Brussels. And I think, what was it 1,800 people subscribed for it. It was Yeah. Wow. Yeah. There’s a lot of interest for the copilot. Yeah, integration for sure. Yeah.
Just to follow up quickly on that, why co-pilot and why not one of the other platforms?
I think Microsoft is everywhere in legal. Yeah, I’ve had I’ve never encountered any law firm, for example, that is using the Google workspace instead of Microsoft. And copilot is basically a new kind of interface and Microsoft decides Word, Outlook, you now also have copilot. So, henchman wants to be aware the lawyer is and that is, that is Microsoft that is copilot. Yeah.
Jorn Vanysacker 15:27
I think the majority of our customers or prospects are experimenting with copilot.
Jeroen Thierens 15:34
Now, it’s crazy. Are you guys experimenting with this? Or are you not able to disclose this on a podcast?
We’re experimenting with everything. Now, whether or not we’re actually implementing something that’s a different story.
Marlene Gebauer 15:52
I’m curious, though, with with copilot. I mean, are you getting any feedback from firms about security concerns? And sort of Microsoft’s ability, you know, perhaps to sort of see content prompts?
Jeroen Thierens 16:07
Yeah, that’s definitely a good question. When it comes to the co pilot, the copilots itself there actually, Microsoft has switched off that abuse monitoring, I think. So there is no security issue when it comes to that or confidentiality issue. And of course, integration with Henchman. Yeah, there we have all the security measures we have in place as a vendor. We are ISO certified and up to date to compliance and we’re constantly monitoring our security. So when it comes to copilots, specifically, yeah, we don’t see any security or confidentiality issue.
Marlene Gebauer 16:48
So Henchman originally a European company got a lot of traction with clients in Europe, for as you said, crossing the pond to the to the US with, you know, clients like you know, Goodwin or Lowenstein. How does the European perspective affect US clients? And also, we’re talking about like US clients, but but also I imagine a lot of these clients have internationally. I mean, they have offices outside of the US. And, you know, is there a difference? I guess, in terms of of approach?
Jeroen Thierens 17:29
Yeah, I think that ties a bit back to the knowledge management perspective, indeed, that in Europe, we’ve gathered a lot of insights on how to build technology that assists in knowledge management. So we’ve built those, yeah, those insights in our product. And we see that a lot of US firms are very interested in that. But also, I think our Europeans perspective is affected is also reflected in the fact that Henchman is a language agnostic system. Just our home country, Belgium, there already three languages just in our own country. So for us, it was really important from day one to build a product that works. And not just in English, but also all the other European languages that are out there. So you also mentioned Goodman, for example, Marlene, their Paris office, for example, is also using henchman because, yeah, it’s works perfectly in French as well. So I think that’s something unique about Henchman, the fact that it doesn’t only work in English, but also in a lot of other languages, and especially the large US firms that are active globally. I’m sure a lot of interest in that specific part of Henchman as well. Yeah.
So I wanted to get back to the the AI buzz which you can’t have a conversation without talking. Yeah, exactly. So how are you guys approaching this? What do you see? Or how do you see AI fitting in to your product? And I really want to kind of think about this, because we’ve had this conversation now for the last 15-16 months. And one of the things that I think a lot of people remember but tend to kind of deprioritize is especially when you’re dealing with with contracts, you want very specific things in your contract. You don’t want a lot of creativity, especially on unmonitored creativity sneaking into the contracts. So where do you where do you see AI fitting into Henchman to keep, you know, the solid thing solid, but to the where it would would be at the most advantage to take take advantage of the AI.
Marlene Gebauer 19:53
How do we keep it precise?
Jorn Vanysacker 19:54
I think it’s I mean, legal legal world has had you know, two quite powerful waves have to say, you know, you had COVID, everyone had to go into the clouds, that was definitely something that, you know, shook the foundations of a lot of firms. And then you had Genie AI ChatGPT arriving. And, of course, it’s something that we and our clients experimented with a lot at the time, there was a lot of uncertainty of what is this going to do a lot of confusion? What will it be able to do? I mean, people thought it was going to change everything, right. And of course, now, we realize now that the dust has settled what the potential and limitations are of generative AI, and similar technologies. And I mean, we see a lot of I mean, in a year from now, we’ll look back, and we’ll see a lot more legal tech startups emerging that have, you know, build something around, like, you know, wrappers around Large Language Models, such as GPT. And if we look at the use case that we serve, the transactional contracting and negotiation, as you mentioned, Greg has to be what it has to be right? It has to be correct. And so the way we kind of, you know, walk, walk people through it is like, let’s say that you, you make a prompt, and you ask it to write a liability clause, right? Then it will generate something, of course, it’s not going to be sufficient, but then some firms are like, but what if we connect it to our database, then it would, you know, play our tone of voice in within that prompt, and they’re probably right, right. And so you get a result that sounds a bit like the firm, it feels like the firm, but within the transactional work, you will always be left insecure of whether you are comfortable using that in this situation, because is this something that my senior partner that I’m writing this for would approve? Is something that we typically use within this type of contracts. And so in that sense, you miss contexts, and you make you miss context in order to make very confident decisions in your work. And so you need that element of contextuality, which Henchman is really good at in terms of finally retrieving that from your DMS, attaching the meta data and kind of funneling that to you when and where you need it. And that’s really how we see our unique position. I mean, it really helped us it was a it was a true blessing for us. Because we now look at Large Language Models in general as a tool to serve, to serve use cases much better and much faster. But it’s not the core of our products, etc. And so we look at use cases all the time. And then we compare different LLM, we have a multi LLM philosophy, so to say, we compare different LLMs to different use cases that we look to solve our customers could be, you know, pluralizing clauses for them, or kind of how can you semantically compare clauses to one another, in let’s say, like a matrix view, right? It’s very good at structuring text. So that’s what we like to use it for. And so in that sense, we, we build use cases, we build out the desired outcomes, and then we measure LLM results against one another. Also taking into account certain specifications such as, you know, risk of hallucination, costs could definitely be one as well speed as well. And those things we constantly take into accounts. And so I think, if you look at Gen AI, if you look at copilots, in general, they’re actually quite reactive in nature, right? You ask it something, and it gives you something. It’s often connected to a data set, which it’s conscious of, but it’s not factual. Right. So it’s very good at providing you with the next best words, so to say. Whereas at henchman, we’re all about going beyond that. We’re not building a copilot, we are building the Chief of Staff for each legal professional. And the Chief of Staff is someone who is very data driven, very strategic, and he or she will, you know, will approach you when he or she thinks it’s necessary. So very proactive in nature. And so, for us, a Chief of Staff is all about providing you with the next best options, rather than providing you with the next best words. And that’s what we’re all about. That’s how we you know, use Gen AI, ai and Large Language Models to kind of contribute to that mission.
Marlene Gebauer 24:30
So this is the part in the podcasts where we ask our guests our crystal ball question. So Jorn and Jeroen, what do you see as a change or challenge that you’ll need to address in Henchman or, or the industry as a whole over the next two to five years?
Jeroen Thierens 24:53
I think when it comes to the industry as a whole I’ll throw a word Gen AI.
Marlene Gebauer 25:01
What’s the what’s the what’s the biggest challenges for free? There’s so many. But if you had to pick one or two, what would those be?
Jeroen Thierens 25:10
Yeah, I can definitely think of something you
Marlene Gebauer 25:13
We could have an entire podcast on that topic.
Jeroen Thierens 25:16
Yeah. Something more specific I thought about when it comes to Gen AI, is the fact that certain Yeah, Gen AI products require a lot of new skills from lawyers. I’m not talking about Henchman. But there’s a lot of Gen AI tools where you need to do a lot of prompting. That’s very interesting, of course, but there’s a whole new skill set needed for that, and that the big majority of people don’t have today. So I think a lot of new skills are needed there. And especially change management when it comes to back that I think people really have to change the way they work, and learn to do prompting. And I think that will be quite a challenge in the coming years. So that’s a specific thing. When it comes to Gen AI in law, that will be an interesting challenge, I think in the coming years.
Jorn, how about you??
Jorn Vanysacker 26:16
Well, I think what I am a big believer of, of knowledge, the concept of knowledge in legal and it has increased significantly in the last year. And I feel like, you know, all the big law firms, all the leading law firms realize that as well, like, it’s their competitive advantage against other law firms. It’s their competitive advantage of getting their people as performant and ramped and quickly ramped as possible. And it’s instilling like a sharing mindset. Within those organizations, which is have obviously it’s a cultural thing, right? You know, it’s a people thing to people, person to person thing. But that can be a real enabler to instill a sharing mindset within those organizations. And I think the law firms that are focusing on that and you know, striving to be successful at instilling a sharing mindset, they will remain to be very successful. And I think knowledge is a very important ingredient in achieving that.
Jeroen Thierens 27:20
Set, I saw Marlene jumped, jumped off and back on and I think the wrong equipment is on. So I get to wrap this thing up. Well, sorry about the technical difficulties at the end there but Jeroen Thierens, Strategic Account Advisor and a Jorn Vanysacker, co founder at Henchman, I want to thank both of you so much for coming on and talking with us at The Geek in Review.
Thank you.
Jorn Vanysacker 27:53
Thanks for having us.
Jeroen Thierens 27:54
All right. And then here I am is Marlene. So and, of course, thanks to all of you for listening to The Geek in Review podcast. If you enjoy the show, please share it with colleague. We’d love to hear from you. So reach out to us we can be found on social media. Marlene is on LinkedIn and on X at @gebauerm and on threads at gay Bauer M six. I’m sorry, @mgebauer66. And I can be reached on LinkedIn or on X at @glambert. Jeroen and Jorn, where’s the best place for people to reach out if they want to learn more?
Yeah, definitely LinkedIn as well. My name is Jeroen Thierens. I’m very active on LinkedIn. Also Henchman as a company is very active on LinkedIn. You should definitely check out Steve Lits. He’s the best guy. He’s our main character on LinkedIn. You should definitely check him out. You can find him on LinkedIn as well.
Yeah, I’ve watched a few of those videos those are those are great. If the audience hasn’t seen the videos good to go check those out those that’s worth the price of admission right there.
Jorn Vanysacker 29:04
Thanks so much for having us guys.
Greg Lambert 29:05
All right. Thanks, guys. Thanks Marlene!