Summary

In this episode of The AAAi Podcast, we spoke with Ryan Samii, Head of the Legal Vertical at Hebbia.ai, about how lawyers can drive meaningful innovation by moving first and learning along the way. Ryan’s career path—from practicing law at Paul Hastings to launching a startup and now helping lead product at Hebbia—reflects a shift happening across the legal industry: one that favors action over analysis paralysis. Instead of waiting for fully formed answers, Ryan and his team have found that building and experimenting is often the fastest way to discover what works. This episode offers insights not only into Hebbia’s approach to AI product development but also into how legal professionals can adopt similar strategies in their own organizations. 

Key Takeaways

1. Start with Action, Then Refine

Instead of waiting for complete certainty or a perfect plan, Ryan emphasized the importance of building and testing early. By putting ideas into motion, teams uncover real-world feedback that helps sharpen strategy more effectively than theory alone.

2. AI That Supports Knowledge Work

Hebbia’s tools go beyond traditional legal tech. They’re designed for knowledge workers across industries, especially those who need to parse large volumes of complex, unstructured information. Legal professionals benefit from this design approach because much of their work involves understanding and organizing dense content.

3. Innovation Beyond Chat Interfaces

While many associate generative AI with chatbot interfaces, Hebbia chose a different path. Their platform offers a structured, grid-based interface that mirrors spreadsheet functionality. This approach allows users to track patterns, extract key language, and build repeatable workflows—without relying on natural language chat as the primary interface. 

Final Thought: A Podcast with Impact

This episode highlights a core message for legal professionals experimenting with AI: insight doesn’t come from standing still. Whether launching pilots inside a firm or testing a new product, progress comes from doing. Ryan Samii’s journey and Hebbia’s evolution show that the most useful information often emerges only after you take the first step. 

Watch Episode 11

Transcript

Below is the full transcript of Episode 11 of The AAAi Podcast.

Bridget McCormack: Welcome to the AAAi Podcast. I'm Bridget McCormick, the CEO and President of the American Arbitration Association. 

Zach Abramowitz: And I'm Zach Abramowitz, founder of Killer Whale Strategies and investor in disruptive legal startups. The AAAi Podcast tells the story of the American Arbitration Association's AI journey. We discuss emerging trends in artificial intelligence with movers and shakers inside the AAA, as well as with key influencers in the broader legal AI ecosystem. 

Zach Abramowitz: So Bridget I told you this story, but last year I was in Legal Innovators, California. I had just been talking about the work that you were doing at the AAA on stage, and this fine gentleman over here comes over to me and. Introduces himself says, I'm Ryan. And tells me, by the way, I went to University of Michigan Law School, Phil Blue, and the way he put it was Professor McCormick was a celebrity.

Bridget McCormack: Should I know about that?

Zach Abramowitz: You should know. I'm telling you. He said that she was a celebrity. And then he said, by the way I'm about to join a legal tech company. I can't tell you now, but I'm going to be able to tell you next week. And then I learned about Hebbia, one of the fastest growing, really most exciting companies in our space that Ryan had joined, not just as a lawyer, but really as, legal tech.

Zach Abramowitz: So I thought, wow, what an incredible opportunity, first of all, to talk to the person. Running the whole legal part of Hebbia because they're not exactly a legal tech company, which we'll talk about. But also, what an opportunity to learn about Bridget McCormick, the celebrity, the professor at university. So I was excited for both. But let's start, Ryan, and tell us a little bit about yourself, your background and how you ended up at Hebbia. 

Ryan Samii: Sure thing. Also, case in point, I feel like all roads in the legal world at some point converge. And it could be from law school into legal tech into practice and kind of mixing and matching between them all. So I guess I'll start from Ann Arbor days. Went to law school, University of Michigan. We can revisit that. And my celebrity comment maybe in a little bit, but graduated went into private practice, was an M&A and emerging growth lawyer for about five years at Paul Hastings in Los Angeles.

Ryan Samii: Really enjoyed a lot of aspects of practicing in M&A. So much of deal work I found to be very intellectually stimulating. But there's a whole other side of deal work as well, which wasn't quite as stimulating. And I think I was in a place where a lot of that compounding competence that you feel in like the early part of your legal career, which is actually like very fun and very enriching, starting to feel some of those diminishing returns. And I wanted that next adventure, that next story and I felt this pull towards starting a company the entrepreneurial journey that I'd always foreseen for myself. So Left Big Law started a company which was in the contract standardization, negotiation, automation space.

Ryan Samii: Effectively trying to take a lot of the high frequency process documents. I saw a lot of my clients, frankly, going to war over for no good reason. Built a company in that space, couldn't have predicted this little product called ChatGPT, that came out shortly thereafter. But that sort of, put me at this really interesting intersection of. Law firm, tech, law firm, technology, in-house legal AI. That journey ultimately brought me to Hebbia and brought me to New York City and the rest is history. 

Zach Abramowitz: And tell us a little bit more about your role at Hebbia, 

Bridget McCormack: Tell us about Hebbia. 

Ryan Samii: Yeah, goes to show like sometimes you take a leap of faith, and you don't really know what the surface area of luck that you're going to create for yourself really is. This is one of those stories. I was hoping for something where, I could draw upon that lawyer background being an M&A lawyer but also start dipping into the lessons I learned as a founder and all of that. So my role at Hebbia now a broader team that I'm a part of has been about bringing what is effectively this very powerful AI platform for extracting, analyzing, and synthesizing information at scale, using ai, large language models to the legal industry. So Hebbia had really gotten its initial foothold, those waves of traction, adoption, and financial services.

Ryan Samii: A lot of the things that people in the finance world want to do with unstructured information. Map to the things that lawyers want to do with unstructured information. And oh, by the way, lawyers tend to have a lot more unstructured information than people in the finance world. So definitely there for Hebbia getting into legal. I came on board to figure that out along with some colleagues. And to your point, we've been moving quite quickly 

Zach Abramowitz: in the last year. And Bridget's question, tell us more about. Hebbia itself. 

Bridget McCormack: Yes. Does Hebbia consider yourself a FinTech company, a legal tech company? Both just an AI company? How do you tell what do you tell the world about what Hebbia is? 

Ryan Samii: Yes. I wonder if we're like part of, in, in part of that like tech curve of AI where you can plug and play any word in front of tech. That solution kind of play in that space. Yes Insurance, tech, FinTech, legal tech, right?. I feel like walking, even the conference halls you feel that a little bit with these solutions See on the TV screen, so I don't think we, we tell this story of we belong under one category. I think one thing that Hebbia did from the early days, let's call the early days, post ChatGPT, even though Hebbia was founded pre-ChatGPT, nearly five years ago now, was we wanted to build the best orchestration layer over ai, large language models.

Ryan Samii: For knowledge workers to be able to operationalize information. So for someone in finance that looks like a certain journey of trying to make an investment decision, working over 8-Ks, 10-Ks, things like that. To learn about an industry for a lawyer that's operating on your M&A agreements, your depositions, your credit agreements, trying to figure out what's market, not just in the public markets, but for your own private documents.

Ryan Samii: So the common denominator is knowledge workers want to be able to synthesize information at scale to make better decisions. That doesn't seem to me to be a finance world thing, a legal world thing. And I think that's the reason we've found success in a lot of different pockets of the market, whether it's legal, finance, real estate, insurance list goes on because building a really effective, powerful tool is something that all those industries can feel the same way.If I put the question back to you, Microsoft Excel. 

Zach Abramowitz: Is it legal tech? 

Ryan Samii: At a certain point, like pinging a SQL database for data. Now, for data analytics, is it legal tech? Is it 

Bridget McCormack: FinTech? Yes. 

Ryan Samii: At a certain point, I think it, it's an interesting challenge to try to build a horizontal solution upon the leading large language models that can service all these different areas of Aldrich. But I think if you do it well and you build something delightful, the lessons you actually can take from the finance world. You find ways to work, bring that into legal and vice versa. So it's actually been a really interesting perch for us to build from. 

Zach Abramowitz: And what's interesting is, on the one hand you say legal is more specific, more niche, going deeper and we're more than that where you mentioned finance, insurance, real estate. The funny thing is about those industries with legal as well, is they're all deeply connected. Meaning to isolate finance from legal or insurance from legal. These are the big purchasers of legal services to begin with. And it almost in a way, makes sense to not think about it as. We always talk about silos within an organization, right? But like silos within commerce more broadly. 

Ryan Samii: Totally agree with that. And I feel like what AI enables probably just like puts underlying double circle, bold, everything you just said because it's even more relevant now because you have this new way of orchestrating information. But information was never supposed to just sit at the law firm. Like that information was always to provide better client service. So to your point, doesn't really make sense to envision that you're in these zone buckets. So that has ramifications. I think in terms of product. It also has some ramifications in terms of company building go to market. It was pretty interesting to come into Hebbia and have some of the biggest clients of the biggest law firms in the world were Hebbiavier users. 

Bridget McCormack: Yes

Ryan Samii: And maybe it was the front of house investors, analysts being like, wow, this is something. Of course my in-house legal team should be able to do these things. Should be using it.. And the in-house legal team's. Of course, like outside counsel should be able to do a thing or two with this. Yes. So that was like an interesting channel and what's been most fascinating is now we're seeing the reverse play out too, where some of the law firms we've brought on board are thinking so why exactly are we adopting ai? We want to provide, better client service and better client service, hopefully has all these other positive externalities. To the extent we're building something really interesting and Hebbia, we're building this deal points library. We want to socialize that with our client. We want to bring our client into this.

Ryan Samii: To your point, if that product begins and ends in a certain vertical you lose out on that. And I think one of the interesting prospects of AI is this the first time a I'll throw air quotes on it, like legal tech product could actually stand for the proposition of you're engaging with your client, with the stakeholders in a end-to-end deal process, in an end-to-end arbitration process are now playing in the same platform could be pretty tricky. 

Zach Abramowitz: And what a flywheel as well, to be able to say finance brings the product to legal. Who brings the product, back to finance in a certain sense. It's an incredible go to market in a way. You have a sort of interesting interface that your users get, that may not be like, I think people think about generative AI. They are picturing like a chat interface, lawyers and finance folks haven't worked in a chat interface forever. They've worked in different ways. Just synthesize information. Talk about like this unique user interface.

Ryan Samii: Yeah. If we're having this conversation like nine months ago talking about the application layer of AI, that was a space that people said you shouldn't actually play in. And the value's accruing at the model layer. And so it's been funny seeing how these things have evolved. I feel like you and I were talking about this in Nashville. Yeah. And now the Hebbia lines are, what we were talking about back then. So I would say what Hebbia did really is we recognize chatbot as a form factor for AI. Makes sense why that was like the first moment in the consumer market. It's it is a good interface for a lot of things. But if you take out ChatGPT and you think about what is a chat bot, what did a chat bot represent to you before, what was it? November, 2022. 

Bridget McCormack: Yep. 

Ryan Samii: It's like where I would put my complaints into Uber Eats or where I would tell Nike get, or I get frustrated with my bank. I got the size nine and a half, like you sent me the wrong size. It wasn't really l an interface for any kind of work, let alone like high-end knowledge work. So the fact that you're putting a user into an interface where they have to have this conversation, they need to iterate on it, they need to rebuild it each time. They want to do something that is otherwise repetitive. Like when you sit down with an M&A agreement, chances are you care about the same deal points, like deal to deal. So that, to do that work in a chat bot didn't make sense. What Hebbia really pioneered was that sort of data grid led interface for large language models. So the proverbial. Microsoft Excel is to structured information. Numbers what he's matrix product became to unstructured information. Text on paper. Right now, that grid interface, I'm not going to lie to you, lawyers didn't go to law school to become like Excel masters. Excel is always like a spooky place for lawyer supply. If you needed to build like a funds flow as a corporate lawyer, you're a nightmare. So from that standpoint, it, it was a little bit different. Lawyers are used to linear document based interfaces. 

Bridget McCormack: Yes. 

Ryan Samii: But what's a grid? Really good for provenance of information, seeing all these discrete, decomposed data points, all of which need to come together to answer that higher order question. You have doing things in a repeat way, iterating on, a discrete moment in a broader end-to-end workflow and building a library of information. That's where Hebbia really made a mark. This often happens in tech waves. You've got to just keep staying ahead of the market. Like data grid is an interface. You walk around these conference hall, you're seeing it, you're seeing it all over the place. So that's where we're continuing to want to evolve and adapt and really think about this data grid playing a role in overall agentic workflows. What we're doing behind the scenes from a technical, different differentiation standpoint.

Ryan Samii: But that grid was certainly something that the legal market liked when they first saw it and they continue to really leverage up to now. 

Zach Abramowitz: And the pace of innovation at Hebbia you were telling me about this, like this is, it's not necessarily the place you want to go if you're looking for a lifestyle like play this is fast moving company. Everyone is in the same office. Talk to us a little about the culture right now of working in one of the fastest moving AI companies. What does that look like on a day to day? 

Ryan Samii: It's fun. I think it's fun operating in a part of the cycle where the same kinds of like headlines that my parents might read any given day. Oh, we're hearing about this thing deep seek. It's yeah. These things are very relevant. Yeah. And like our line of work, like these are the lunch table conversations we're having. These are like the ways we're iterating. Yes. Hebbia we're a very fast moving company. It's one of those hard things to continue as you're on your scale journey. But. We're down in soho five days a week in office and we really pride ourselves on wanting to continue being opinionated, being contrarian in like some of these decisions we make. To continue really advancing the ball forward for certain verticals that are not easy to solve problems in, finance, legal, these are like messy verticals, so you have to make a lot of thoughtful decisions. The way you make thoughtful decisions is sometimes by. Moving quickly. Like I think one of the mantras that certainly our team, the legal vertical has rallied around is as a lawyer so much training in what you do, you take action on the basis of a lot of information. 

Ryan Samii: Before you make that decision in drafting a motion, before you make that decision and marking up a contract, you're looking at your precedent. It's like information produces action. The thing I think you never want to lose out on when you're. Scaling a company is actually a lot of action. Supplies you with information. It's like the flip by being fast moving and continuing to experiment and iterate. You find information comes out of that because if you just sat back and you tried to make assumptions as to how will this enterprise think about this kind of form factor of AI. You're not actually going to be able to make that decision, reach that conclusion. So being fast moving allows us to get into those cycles really quickly, make some conclusions, move on to the next experiment, and we've done a good job retaining that energy. 

Bridget McCormack: How big is the team and how many other lawyers are there on your team?

Ryan Samii: I think we just recently hit the triple digit mark as a company. 

Bridget McCormack: Oh, wow. Yes. 

Ryan Samii: Growing in terms of what we're doing with the legal vertical, I would say we have really grown that function with subject matter experts coming in. People who've previously been in legal technology, people who've been really thoughtful about change management at a major enterprise, that's not the kind of skill you necessarily get just by, having 10,000 hours you put into negotiating credit agreements. That's like a different kind of challenge. So we've grown the team in that way. So I would say generally the people at Hebbia that are really thinking about legal as this interesting opportunities that we want to continue building in. We have quite a few folks now, like probably. And like to the earlier point, a lot of things we're doing on the product engineering side of the house. We want to make these really thoughtful decisions where the benefits are felt in multiple verticals. That mean there are moments where we have to be really thoughtful about a lawyer's workflow and how that is perhaps different than someone else?

Ryan Samii: Of course. But that's where we have a lot of people at he that either exclusively or at least in part, are really thinking about legal and the opportunity there. 

Zach Abramowitz: I want to go back to what you were talking about before with action leading to information and the benefits of moving fast. I, and again, I think so much of the AAA's strategy on AI was exactly that. When you came in, you were not every other law firm and legal organization, anyone in our industry wanted, I. Subcommittees and Task forces. Task forces, and it's this was, you said, go to die, said none of that. You said, let's just start building. And there have been a number of projects. If you sit in on the AI steering committee meetings at the AAA, you will see constantly new projects, new developments.

Zach Abramowitz: And yes, there was a rigorous criteria to decide what we're going to build. But you didn't let it get to the point where it was like perfect is the enemy of the good. And not only that, by doing multiple projects. because I think a lot of people have said, start with one. And you're like, no, like we're going to build a lot because by building we're going to develop that AI muscle. We're Action 

Bridget McCormack: Information is was how we learn. 

Zach Abramowitz: Action will learn information. I love it. 

Bridget McCormack: We have to make some bigger bets. I didn't want to make the bigger bets until I had a lot more information. The only way to get the information was to like move fast in lots of directions. So I completely agree with that.

Bridget McCormack: I think it's like the best way to get information 

Ryan Samii: and I think the enterprises are feeling this too. The law firms that sort of made that uncomfortable decision and maybe some of the early innings of ai, they're just like a different part of this cycle now. They are. And how they interpret new solutions, evolving solutions. Starting now with the hope that everyone else would've led you to the right place. Doesn't look like that's actually going to be the right strategy. because you haven't found information for your organization. The only way to get there is by taking decisive action. But being comfortable, going back to first principles, like what's we learn from this? Now we actually have this information, now I make the next decision accordingly. 

Bridget McCormack: And the change management piece, I don't think all legal organizations like change management is really hard, super hard. And if you hadn't started that two years ago, starting it now puts you deeply in the hole. I'm not saying you can't find a way to catch up, but I'm glad not to be in that hole. 

Zach Abramowitz: And Jae Um said a few years ago when it came to legal technology the risks of early adoption outweighed the benefits at the time, but there was a point where the, where that switch became the risks of not adopting were greater than the benefits of waiting. 

Bridget McCormack: Yes, she identifies when that flipped. I forget what she says. It wasn't the moment 

Zach Abramowitz: of ChatGPT it was slightly before, if I recall correctly, but I think, to that point, action leading to information was also your personal journey because you made a very clear decision to leave the law firm, start a legal technology company at a point that you couldn't have possibly predicted was going to let you know it wasn't the right timing. And by the way, I invested in companies that got caught in that same exact issue, right? Where it was just a matter of timing. And if it had been, if they had started three years earlier or later. It would've been great. But what it led to for you is information because you understood, okay, everything I was trying to build was like extraordinarily difficult. Now with AI, It's like table stakes. I can go much farther. And it just goes to that point about the compounding interest of early adoption. In your case, it didn't end up in a billion dollar exit, but it did end up in you joining one of the fastest moving AI companies. Very early on, which ultimately may end up being like, a path, to generational wealth for yourself as well. So action leading to information. I love that. 

Ryan Samii: Yes. I think also just like your personal journey, there's like a version of it. I feel like a parallel to that, and this is a lesson, my dad always been parted with me. You can't necessarily control the moments where luck finds you.

Bridget McCormack: Yes 

Ryan Samii: But, maybe you can increase the surface area of luck. 

Bridget McCormack: Yes. 

Ryan Samii: And it could be the case that, if you're there doing your hundred 73rd SPA markup, you're in a different surface area of luck then if you're chatting with Chief Innovation officers about their AI strategy and you're chatting with head of legal operations of private equity firms about what they're seeing in the market, these things are so hard to predict. It's like the same challenge of adopting AI, not really knowing where that's going to take you, take that first step and interesting things can happen from there. And it's action producing information. 

Zach Abramowitz: Yes. Paul Graham has made this point many times that one of the best ways to get rich is to start a startup, but not necessarily because that startup will hit, but because if you are starting a startup and you're in that ecosystem. You may end up having an opportunity to invest in another great startup, to join another great startup. Maybe you learned about Bitcoin early on because you were in the sort of tech ecosystems. 

Bridget McCormack: No, it's the ecosystem is the benefit, right? 

Zach Abramowitz: Yes, exactly. 

Bridget McCormack: It's an unusual move though for there aren't that many people who go to big law and then are willing to step out and start their own company. Like what do you think explains that? What is your secret sauce? Because in fact, I think a lot of people that go to law school are exactly the opposite of people who are willing to go found a company. So tell me what is Ryan's secret sauce. 

Ryan Samii: Yes. The, I wish it could be like the nutritional label that I turn around here and it's like you're, the ingredients, like the secret sauce. Like you mix it up. I think it was like a series of different things. I'll come back to one phrase you Hebbia a lot, I think like at a conference like this, is like recovering lawyer. This idea that, yeah, I did this thing now I want to that, I actually don't love that. I don't think being a lawyer was something I want to recover from. In fact, I look back on it very fondly. I think for me, there's a version of a career that I saw I could become increasingly more expert at this nuanced area in emerging growth or M&A. And that's totally fine. I think becoming like a real bonafide expert top of your field in something is like journey that some people really enjoy. Yes. I think for me the zero to one journey of I'm tackling something new, I have to learn to then apply, which by the way, I felt during a lot of those early days going from one deal to the next deal. Yes. You're sharpening up your tools, you're bringing into the, as soon as I felt that a little bit less, I just wanted to get onto that journey again. It was that sort of personal,

Zach Abramowitz: Once you hit the sort of plateau, you're like, like now I want to keep looking for the summit.

Ryan Samii: Yes and it just gets to a point where every morning you wake up and you just feel that desire and you have enough of those days back to back and you're like, all right. Time to make that decision. And for me too. Yes. I thought I perceived of a problem that was solvable and worth solving and, yes the looming cloud of not attempting something to me felt worse than attempting it and wherever that path takes. 

Bridget McCormack: It’s amazing. 

Zach Abramowitz: All right, so we've talked Hebbia, we have talked about your personal journey. Tell us about Professor McCormick. The celebrity of Ann Arbor. 

Bridget McCormack: I thought you were going to ask, is Michigan going to beat Auburn on Friday night?

Zach Abramowitz: Who needs to ask? 

Bridget McCormack: Okay. 

Ryan Samii: Six straight. Sweet. 16, I don't know maybe a blue blood basketball program in addition to football. 

Bridget McCormack: Yes. 

Ryan Samii: Yeah, Ann Arbor, Michigan law it's a very tight knit community, but I think there's something about this probably at every law school where as a student, you're sitting there and maybe celebrity isn't the right word, but there's certain people in the faculty that are these larger than life figures for whatever reasons, right? If I have any piece of advice to a law student, I think the mistake that I made is I was thinking, I'm sitting there, I'm a three out at Michigan Law. No, I want to go be a transactional lawyer. So I was like, I've gotta do the transactional courses. I need to do M&A. The biggest piece of advice is if there's anything cool you can do, just go and do the cool thing. So I remember I had a couple really close friends who were sitting in her class. I think it was State Supreme Court practice or something like that. And that was such a cool class I should have sat for. But I did have the benefit of, lunch sessions and things like that, where if you're not going to the lunch session where Zingerman sandwiches are being served, then you're going to a lunch session where there's a, one of those on campus celebrities. Yes, that's where our paths cross back in Ann Arbor. 

Bridget McCormack: Very kind and fair to me. What's the next one? 

Zach Abramowitz: There is a deep recommendation and I hope like that the more senior attorneys who know people who are going to law school right now will pass this advice along. Because I think the best thing that I did at law school was when there was a personality like that. And as an example, Evan Chesler, the managing partner at Cravath, taught a course on trial practice. That was awesome. That was one of the best courses, best experiences I had in law school. And there were other experiences like that. And law schools do have these people. If you're in law school don't think about witch doctrinal. Just go you're going to, when you do the law go towards the awesome. Yes, totally. You learn on the job. And I think the simple rule is if it's cool. You do it. 

Ryan Samii: There's another professor in Michigan, law Professor JJ White. Do you know a book JJ White wrote? I don't. The Uniform Commercial Code? The UCC. Oh, wow. I didn't even know that. That was a book that anyone wrote. I thought it wrote of those came down from the sky, from Sinai. Yeah. You told me they wrote this in the Greek city states and it was passed on through Yeah, that, that makes sense. Nope. There was a professor in Michigan law who was, the driving force of the UCC. Go sit in his class, if you're in law school.

Bridget McCormack: Yes I would say the, but you have to be willing to be there at 8:00 AM because he only teaches contracts at eight. That's right. He's a Marine. 

Zach Abramowitz: That's awesome. Yes. Ryan, thank you so much for coming and sharing a bit about your story and we're going to be following, have you closely. 

Bridget McCormack: Yes, we can't wait to watch what you do. 

Ryan Samii: Fantastic. Yes. Thank you both. Really enjoyed it. 

Zach Abramowitz: Thanks for tuning into today's episode of The AAAi Podcast. Please consider leaving us a rating and review on your favorite podcasting platform, and don't forget to follow and subscribe so that you never miss an episode.