From Copilot to Colleague: AI at Thomson Reuters with CTO Joel Hron

Hi there and welcome to Enterprise AI Innovators, a show where top technology executives share how AI is transforming the enterprise. In each episode, guests uncover the real world applications of AI, from improving products and optimizing operations to redefining the customer experience. I'm Evan Reiser, the founder and CEO of Abnormal AI. And I'm Sam Motamity, a general partner at Greylock Partners. Today I'm speaking with Joel Rahn, CTO at Thomson Reuters. Thomson Reuters is a global information services company with more than 150 years of history. They're the world's largest legal software provider and are home to some of the largest tax and compliance software businesses in the world. Their perspective on AI is especially compelling because they're deploying it directly into the knowledge work that lawyers, accountants and tax professionals depend on. Every day. A few things stuck with me from this conversation. First, Joel drew a clear line between using AI as a co pilot versus using it as a colleague. The shift happened when one of his engineers noticed that 50% of their code was being written by AI. At that point, the engineer isn't the primary contributor anymore, they're the controller. Second, Thomson Reuters has built agents that take a stack of tax documents, extract the data, apply the tax law and compute a complete tax return. The tax professional's job shifts from pulling numbers off documents to advising clients on their financial decisions. AI isn't just improving a workflow, it's filling a real labor gap. And finally, AI has made Joel more technically engaged as a CTO. With 5,000 engineers and 100 plus products, it was impossible for him to ever look at anyone's code. Now, before a Friday engineering meeting, he can clone the repo and spend an hour with Claude or Codex to actually understand what's happening. Joel, thank you so much for joining us today. Maybe kick us off, do you mind sharing a little bit about kind of your background, how you got to where you are today and your current role at Thompson Reuters? So I've been at Thompson Reuters for about four years now. I've been in this role as CTO for just over a year. You know, early in my career I actually didn't start out in software. I started out in mechanical engineering. I was actually working on simulators and sonar systems and I worked in the energy industry for a while. And I think those early days is really what kind of like got me into computer science and linear algebra and sort of by translation, AI and machine learning over the course of years. So I did that for quite a while in my career and at One point timing felt right to go out and help start and grow an AI startup that we did with, with a few guys and grew that over the course of five or six years and successfully exited to Thomson Reuters. And that was about four years ago. And so, you know, when I joined the company, it was a really interesting time. You know, obviously the first days were really focused on taking what was a small company of maybe like 80 or 90 people and really integrating it into the fabric of the larger enterprise. But like it was a really interesting moment about nine months after I joined and OpenAI released ChatGPT and for a company like Thomson Reuters that works in law and tax, it was like quite a. And as it was for every company, regardless of industry, quite an abrupt moment of like, okay, what just happened? What does this mean for me? How do I need to reflect and adapt? Like our strategy top to bottom. And having just been acquired not so long ago and having quite a lot of experience in the applied AI space, yeah, I found myself kind of front and center in that conversation of trying to help shape our strategy and in the early days of how we adapted to those changes. So that was really phenomenally like interesting time. And I think it's really just kind of grown and blossomed since then and it's been fun to really operate at that scale. Probably no one really fully appreciates the full scope and scale of operations. You might share a little bit more there just so people have a bit more context about kind of you and the organization. So the company has existed for more than 150 years, dating back to like Reuters used to deliver the news on carrier pigeon, kind of old. So quite a tremendously rich history, but also a rich history that's sort of grounded in some really, I think time tested, but also perpetual kind of ethics and truths. And mainly that's around delivering trusted information to the world. And so as you think about that job to be done, certainly most people recognize the Reuters news business, but you know, we're the largest legal software provider in the world. Probably most famously known and recognized for our Westlaw product, which is the number one legal research product in the world, but also practical law, which is the largest legal know how content set in the world. We also have some of the largest tax products in the world, both in terms of we sell primarily, not like a household name like Intuit necessarily, but we sell to large accounting firms, large and small accounting firms primarily. But we also have quite a tremendous like transactional tax compliance business where we like anytime you might go online and buy Something and the tax gets computed. Like that's like our, our systems doing that work. And so we, we feel like a big part of the supply chain from that perspective. And also a large like compliance and audit software business as well. Again like large, most of the large audit providers in the world license software from us. So we kind of touch a lot of segments of the market that really anchor on this, like deep knowledge work and deep expertise, like being a pivotal part of it. And most of our products have some element of like that expertise and content that power what they do. And that could be like a tax engine. It could be in the case of legal research, like a bunch of case law and markup. But we're also one of the largest employers of legal professionals and tax professionals in the world. And so we have experts in house that support the development of a lot of those underlying content assets and data assets that drive the products at the end of the day too. So yeah, that's a little bit about who we are. Like I said, I definitely didn't appreciate or understand the sort of breadth and scale of that when I first joined the company either. But it's been quite eye opening getting to know that. What do you think is at least going to be one of the big jumps in terms of how most enterprises will get the full value of applying even the current technology today? I think from an enterprise perspective there's really two things that I think today are maybe holding back the pace of the adoption. One is sort of what I mentioned before, which is trust. Like I think not just from a like accountability of the accuracy of the AI system. Like that's one element of trust, but also the element of, you know, what do I explicitly trust an agent to access in terms of like well, what files can IT access on my computer, what other web applications can it access, what can it access? My password manager. And if it does access those things, what can it do with those things? And like the control system around the agent is something that I think really needs to be built out further for enterprises to really get comfortable with allowing agents to just do work in a more independent way. And then the second element of trust, like I said, is like building the processes around agents actually doing work versus just being thought partners. And I think that's kind of like the step that needs to happen from a human change management standpoint. And I think you're seeing this happen in software engineering right now where really the job is being redefined and people are sort of now formulating a different understanding of what is My role as an engineer and the value that you create as an engineer is not the lines of code that you write. And in fact, it's never been the lines of code you write. The lines of code were just the way for you to distribute and express your ideas and your judgments. And so what being an engineer means is actually changing. And some engineers are jumping headfirst into that. Some of them are like they saw their value as the lines of code and you got to really reframe their thinking there. And I think, I think that kind of reframing of roles will happen more broadly beyond software engineer as we go through the next few years too. Joel, when I was doing some of my research episode, I think I saw something you presented or wrote around you called the transition between co pilot to colleague. Can you share a little more about that? It feels like quite relevant to this conversation. Yeah, I think so. And I think it speaks to this sort of mindset change. And one of my engineers said something really insightful, kind of at the tail end of last year when the hockey stick of engineering capability was happening with agents. He said something really changed for my team when we hit 50 plus percent of our code being written by AI. And it was insightful because it signaled that no longer was a human the primary controller and contributor to the, to the code base, like some other entity was in this case AI. And so now your job as an engineer shifted towards being like the contributor and owner of the code base, to more being like the controller and governor of the code base. And you had to like now worry about building systems around AI to steer and correct and guide what it did, rather than doing that work yourself. And I think the more that we start to think about AI as like a human contributor, that actually like the better frame of mind we put ourselves in in terms of how to actually use it. And I think that's really what I was trying to get to in terms of like copilot to colleague. In some, in some ways, like a copilot is like, you know, you're sitting there and you're just like talking to, to something to like stir up ideas or fill up a blank page. But when you think about a colleague, you actually think about like delegating work. And if I were to go delegate work to a direct report or somebody in my org, I would give them all the information they needed to be successful, but not too much information that I would just do the job myself. And that's the same sort of act that needs to happen with AI, is building Systems that give AI enough information, but not so much information that you could just do the job yourself if you did. And I think that's the sort of mental model that people need to get in in terms of how they should think about working with AI in the future. I think I saw like your, your co counsel product had like a million users. I was kind of curious, like what are some of the ways that maybe your customers or kind of your users are getting? Like how is AI changing how they engage with the company or how they use the products? Like what are some things that maybe some of our listeners might find surprising about that? Yeah, that's interesting. So you're right. A million users of co counsel today and you know we see really like strong adoption. I think a lot of people have looked particularly at the legal industry for a lot of years as kind this laggard industry that was slow to adopt technology. And I'll say like from day one of ChatGPT starting to shape the changes in technology, the legal industry has been sort of like eager and ready to go in terms of trying new technology. And there's been a lot of like really I think eager adoption by the legal industry but also at the same time like a lot of hesitation and you see a lot in the news about various sort of challenges in terms of being wrong in the legal industry, particularly when using AI and the risks of doing that. So there's a lot of apprehension at the same time, but also a lot of recognition that look, we need to lean in here because it's really going to reshape how we do work. So I think that's been really good. Our customers have been really great partners in terms of how we build Thompson. Reuters as a whole has really, really I think invested and excelled for many, many years in the legal research domains, particularly like litigation in this kind of work. And probably about mid year last year we released something called Westlaw Deep Research which you know, probably across all of the legal industry might have been one of the more profound product releases I think in the industry. And we just, we hear like substantial feedback from customers about the quality of the research that's being being done with that product. And just perhaps like you experience using like deep research on the web and sort of like shocked at how it can like read through 200 websites and synthesize that into like a coherent study. Like we're doing the same things with like deep legal research and it's pretty, it's really impressive and it's really I think changed the narrative of how people think about legal research. And yeah, we're excited. We're in beta with our next version of co Counsel right now, which really takes like a similar paradigm to a broader swath of legal workflows. You know, in tax, like, we've actually built products that really end to end, automate the development of tax returns, you know, to take like a shoebox of W2s, 1099s and put it through all the tax regulation and tax law and compute a tax return at the end of the day. And we're certainly trying to scale up the complexity of those use cases, but we see real potential for agents to complete end to end tax returns now as well, and quite excited about continuing to build that product where meet some of the surprising ways that you guys are using AI across the products that like the average person may be unfamiliar with your domain, might not fully appreciate the true kind of impact and power or the kind of new opportunity AI is kind of bringing for your customers. Let me give you perhaps maybe a couple things that would stand out. So as an individual in the United States, you have to file tax returns every year. And you know, for, for many people you might have a, a CPA that does that return. And like, what your CPA would normally do is they would go send you probably, if you're anything like me, like 30 emails to try to get you to give them all the documents that they might need, like the W2s or your 1099s or whatever. And they would sit there and they would like pour through them one by one and like write down all these values into Excel and use their knowledge of the tax law to then figure out, okay, do I take this deduction or how do I do this and map that into a tax calculator and ultimately bring you back a return. You say, okay, well, here you go, Evan, this is what I did. This is what we're going to file, this is what you're going to owe, or this is what credit you're going to get or refund you're going to get and so forth. And with AI today, we now have agents that can do all parts of that process. And so the tax professional now basically takes these documents and puts them into our system. Our system reads them, it extracts all the appropriate data values, it looks at the tax law and makes interpretations as to how certain data elements should be treated based on the tax law, what positions are most advantaged to take in terms of deductions versus not, and then computes all of that into like an actual Tax return. So the tax professional's job now turns much more into advisory. Rather than just doing grunt work of pulling data off of documents and dumping it into a tax calculator, they're advising clients much, much more directly now on how they should think about their taxes or what alternatives they could think about next year in terms of how they could think about their businesses or their personal lives. And so their job hopefully is much more fulfilling in that way. And there's also a significant decline in the number of people graduating and taking the cpa. And so there's more people filing taxes and there's less people doing them. And you have this real constraint in the market. And so AI is really helping fill that void. And I think if you looked across the industries that we serve, there's similar examples of that as well. What about kind of like on the internal AI use run kind of AI adoption? Are there some ways you and the team are using AI that actually, let's start with the team. Back to you. But like, somebody's like, your team is kind of AI that maybe some of your peers find surprising or might be like, oh, wow, we should be doing that too. Yeah. In terms of internal AI adoption, like I said, you know, one of my engineers mentioned this like 50% figure to me and like, I hung on that pretty hard because last year most of our effort was like, all about adoption. It was like, hey, I just want to make sure that I've got, you know, 90 plus percent of my org using these tools every day. And then we shifted and we said, okay, we want to start thinking about not just using the tools, but how we're using the tools. And for me, in software engineering in particular, I think that meant actually shipping code with AI, like making PRs with AI independently rather than having humans sort of curated every step of the way. So that's one of the main okrs that we have like through the middle of the year is to be API first in our engineering teams and how we at least make PRs. I think we still have quite heavy levels of like, human review and these kind of things, but we really want to be quite intentional about how we're using AI first in the development flow, I think in partnership with like, our product and design organizations. I think there's a lot to be said about just being able to create prototypes faster, but, like, at the end of the day, it's not being able to create a prototype. It's that the prototype that the product or design person makes is a much better translation of what is actually in their head as a requirement rather than like a written requirements documentation is like. And you can actually use that artifact to then develop with AI to say, I need it to do this, it needs to look like this, like, and you're actually using this mock up as sort of the requirement spec to AI. And there's a much tighter loop there in terms of things getting lost in translation and lack of communication. So those two things together I think have a really powerful flywheel effect, if you will. I think the last thing that I would say is really, I'm really still trying to get better at this at the moment, but it's really about bringing better transparency and visibility to the whole organization about all the things happening in the organization. So like I believe any engineer on my team should be able to ask like what's happening with our customers? What are, what did customers say about this product feature in the last, you know, 30 days? And we should be able to look at like every interaction we had through support or other mechanisms to help give our engineering teams more directly a picture of like what is happening with the state of their application or the state of this product scope or whatever it may be. And that's just one example. But I think it exists in sales, it exists in marketing and I think this sort of like data layer of AI is like incredibly important and empowering for the rest of the organization. And I think, you know, we're trying to spend a lot of time getting, getting that right and making it available and accessible to more people. And they can already see in examples where we have done it. It's really opened up a lot of doors. When you think about kind of your team and you're thinking about building the next generation of a AI native or a fully AI powered team where yeah, what are some of the skills that kind of go up in value versus down in value? Yeah, 100%. I actually wrote an article on this recently too, but I think, I like, I really think that, that these deep specializations in some areas become less important. I don't think that they're unimportant. I get, I still believe in sort of this T shaped model of an engineer, but I think the horizontal slice of the tea is more important now than the vertical slice of the tea. And your ability to traverse that horizontal slice of the tea I think is even more important than it ever was before. A lot of people use the term curiosity and I do like that term curiosity. But it's not just about being able to be curious. It's also about being able to learn quickly and fail quickly. So a lot of people can be curious and they go try a lot of things, but they can ultimately just like swim in circles and get nowhere. Like the people who are curious and can apply that curiosity to like learn and evolve, I think are the ones who really like have outsized potential in this world. And I think those are the ones that we really see, I think thriving. I think they're also the ones who like, you know, just personally don't see themselves in a box. Like I told you, I got my academic career was a mechanical engineer. I had no business working in software engineer. But I don't know, I just always identified myself as an engineer and an engineer's jobs to solve problems and like, I don't care what like mechanism I'm using to solve that problem, I'm going to go try to solve it. And I think the more that people can identify themselves as like, like problem solvers rather than like front end engineers or back end engineers or whatever, then like the more they open those doors up for themselves to kind of like do things that they perhaps weren't necessarily trained to do in the, in the traditional sense of the word. Okay, so speaking of new things, tell me a little bit about how like your personal day to day has changed in the age of AI, right? Are there tools they're using, are there workflows have kind of dramatically changed? Are there things that maybe now given you can do so much with AI, become more valuable to spend more time on or kind of, you know, vice versa, like you know, what are, yeah. How are you kind of personally as a leader and executive kind of changing how you work? When I have like ideas that are perhaps like strategic or need to be thought through, like I'll, I'll run like multiple day long convers or multi day long conversations with AI to like work through those problems before I ever bring it to anybody else on my team. Like I try to like test corners of things that are in my head or intuitions that I have quite rigorously before I like talk to anybody else about them. And I find it really helpful to do that just, just in terms of thinking. And then the last thing which is maybe the most impactful is it's allowed me to be like way more hands on than I could have ever been before. And what I mean by that is like, you know, there's 5,000 engineers in my organization. Like it was impossible for me to ever like look at a line of code of any of the people that what they were doing and it's 100 plus products. It's like. But now like I can, if I've got a meeting with a team coming up on Friday and we're talking about a product like can literally go clone the repo and spend an hour like with Claude or with Codex, like talking about the code and what's happening and what happened recently and get grounded in what's going on from a technical perspective and be a lot more useful in that conversation versus just coming in and like talking about the niceties of how the project's going. And that has changed. Like I think how I operate. First of all, it's a lot more like exciting for me to be able to like work at that level. But it also I think actually helps me give better guidance and direction to the teams as well, which is the most important thing at the end of the day, at the end of the show, like do a bit of a lightning round where I basically try to get your one tweet version to maybe some questions that are a little hard to answer in one tweet. So how do you think companies should measure success of a CTO in kind of the age of AI? I think companies should measure success of a CTO based on velocity and impact. Right now. First of all, I think like velocity is just like a non negotiable but, but I think a lot of people are doing like really simple thin things that like have a three month long shelf life. Like I think CTO should be shipping things that like have primitives that are going to be durable. And like I think, I think that will, will prove to be more impactful than just like the flashy next thing. Let's say like you were talking to someone just stepping into a new CTO role and they're like, hey Josh, give me like three must do things in my first three months, six months to make sure that I'm kind of getting, I'm getting the ball rolling and cover owning AI transformation. What would be maybe those like three areas to focus on. First would be on coding tools. I would say like pick a coding tool and get to get to more than 50% of your code being shipped by AI. That'd be the first thing. The second would be talent. I would index on what you believe is important for talent and going forward versus in the past and where you are relative to that. And you know, chart a path to, you know, make sure your talent is mapped to where you need to be going. And then I'd think about what it is you're doing. Like, I think AI has sort of afforded people a lot of opportunities to do it, anything they could dream up. I think it's more important to figure out like what's really going to matter going forward and make sure you're working on the right things. What's your advice for some of your peers about how to make sure they're staying kind of closer to or closer to the frontier of technology, given it's kind of moving so fast these days? Use it. I mean, that's it. Like it just every day just use it. And like, like when you see a press release drop on Twitter or something like that, like go see what it's all about. At least for me, that's how I learn. I just go use it. That resonates with me. I have to play around with it to really understand. Okay, so maybe here's the personal side. What's a book you've read that's had a big impact on you and love to hear why it doesn't have to be work related? I've given this answer before, but I'd say Thinking Fast and Slow by Daniel Kahneman, probably the most impactful book I've ever read because. Well, first of all I like like social psychology, but it also like really helped me understand how, how to work with others. Not just operate and communicate from my perspective and with my bias, but also to like perceive how others might come to this conversation with their perspective, with their bias and like kind of adapt myself to the situation to ultimately like either sell my idea or get alignment or whatever it is I need to do. Like it helps me move faster by having that, I think social awareness, if you will, as well as personal awareness. Because I have my own biases and things like this. And I think that just like recognition and also he presents it in like an exceptionally data driven way I think has just really like shaped how I think about interaction in general. The last question was my favorite one. What do you think will be true about the future of AI's impact on the world that most people consider science fiction today, I would believe that AI is going to go through a bubble. Even though I think it's going to be hugely impactful in the long run. I think as hugely beneficial as I find it to myself today, I think the economic models of how this actually scales aren't quite proven yet. And I think there's a lot of really high leveraged investment sort of betting on it and it feels quite aggressive from that perspective. And so I think kind of like all things. I think there's going to be some short term, you know, bumps in the road as we kind of like figure out how the model of that works. But I think, like, the end state of it will be as impactful as people project it will be. Well, Joel, I got 100 more questions for our version two of this episode, but I really appreciate you taking time to join us. Thank you so much. Yeah, thanks for having me, Evan. Nice to be here. That was Joel Rahn, CTO at Thomson Reuters. Thanks for listening to Enterprise AI Innovators. I'm Sam Motamity, the general partner at Greylock Partners. And I'm Evan Reiser, the founder and CEO of Abnormal AI. Please be sure to subscribe so you never miss an episode. Learn more about Enterprise AI transformation at enterprisesoftware blog. This show is produced by Abnormal Studios. We'll see you next time.