Skip to content
YC Root AccessYC Root Access

This Is The Next Industry AI Will Disrupt

AI is already transforming entire professions like software engineering and law. And accounting might be next. In this episode of The Breakdown, YC’s Tom Blomfield and David Lieb sat down with Onshore founder Dominic Vitucci to find out just how AI is fundamentally changing one of the world’s oldest professions and what that could mean for the future of white collar work.

Dominic VitucciguestTom BlomfieldhostDavid Liebhost
Mar 7, 202633mWatch on YouTube ↗

EVERY SPOKEN WORD

  1. 0:00 – 0:58

    Why Big Four dominance is vulnerable to AI-driven disintermediation

    1. DV

      From my perspective, I understand why it's hard to imagine a post Big Four, a post accounting firm world. It is a foreign concept. Conversely, I don't actually think they've earned the right to maintain the reverence that they've, they've been granted for all of these decades, centuries sometimes. And so if this keeps going the way that I think it's gonna keep going, the way that I'm kind of betting the house that it's gonna keep going, is that's exactly what we're looking at, right? A, a functional and tectonic shift in the way that these outcomes are ultimately provided to customers. Accountants have kind of just wormed their way into the middle and artificially sel- set themselves up to be, you know, "I'm the middleman. I'm the expert." But like what if we just didn't have to have that?

    2. TB

      [upbeat music] Welcome back to another episode of The Breakdown. Today we are joined by Dominic Vitucci, the CEO and founder of Onshore, uh, which uses AI to automate corporate accounting and tax. Dominic, thanks for joining us.

    3. DV

      Thanks so much.

  2. 0:58 – 1:05

    What accountants actually do day-to-day: the spreadsheet labor reality

    1. TB

      What is the job of an accountant and an auditor at a, at a level of, you know, how a c- c- computer scientist might-

    2. DV

      Right

    3. TB

      ... like describe it?

  3. 1:05 – 2:37

    R&D tax credit case study: interviews, rubber-stamped estimates, and manual write-up

    1. DV

      You know, you start off as a junior guy. You're, you're 22 years old. You come out, uh, you, you... Hopefully you passed your CPA exam. If not, you can take it while you work. There's a bunch of professional stuff. But you show up and, and really blocking and tackling nuts and bolts is open a spreadsheet and copy some data from one type of document to a different type of document, and like do some formulas to do some arithmetic, right? So a good example of this. I was doing a R&D, research and development tax credits, right? And it was, I started on Tuesday and on Friday, the partner said, "Hey, let's go out to this client site." It was a client that made boxes. We went out there and it was me, this manager, and this partner. And I brought a yellow legal pad and a pen. And the partner said, "Hey, listen. Don't say anything, just take notes." Okay, easy enough. So we show up. Partner, we have a meeting scheduled all day Friday, and then Monday, Tuesday the next week. Eight hours, 30 minutes with as many employees as we can get at that company. Just like you, Tom, you come sit down and say, "Hey, listen. Here's a piece of paper. Read the paper." And you took, you know, 90 seconds to read this paper. It was all about the R&D tax credit rules. And so you read the paper, and then you look up, and the partner looks you in the eye and says, "Well, how much of your time do you spend doing an R&D tax credit?" And now you're an engineer, you're a sharp guy. You say, "I don't know, 30%. Can I leave now?" Partner says, "Oh, 30%. Uh, don't you mean like 80%?" And the guy goes, "Yeah, 80%." Partner says, "Oh, perfect. Dominic, write that down." And so eight zero, and I did that for 20 something hours. Then the worst part is I go back to the office, open my legal pad, and then I just, you know, Tom Blomfield eight zero equals this times it, and that, that's it.

  4. 2:37 – 4:18

    The real bottleneck: substantiation and proof, not math

    1. DL

      Okay, so the idea of the tax credit is like there's certain kinds of work-

    2. DV

      Right

    3. DL

      ... where the, if you're doing that work, the federal government wants to incentivize it, so we'll-

    4. DV

      Right

    5. DL

      ... kind of give you money back-

    6. DV

      Exactly

    7. DL

      ... if, if you have engineers doing r- research basically.

    8. DV

      Anybody... I mean, in, in this particular domain, and this is one of our like flagship products, right? So I, I do wanna take just a beat to talk about it. Really, it's an incentive at the federal and the state level in the United States, amongst other countries in the world, that encourages research to be done onshore domestically at some of these, uh, organizations, right? The idea being, hey, if you're conducting research and that's the creation of a new product, a new process, or a new technique, or the improvement of a product, process, or technique, you should get an incentive, especially if you're keeping technical jobs in the United States, for example.

    9. DL

      Yeah. What you're basically saying is, like, this, this is a pretty straightforward task of, like, identifying how much of each employee's work is in a sort of area that's, um, eligible for this tax credit.

    10. DV

      Right. So straightforward arithmetically, yes. It's, it's numbers and ma- Like I could... If you gave me a pen, I could just write down how to do the math right now. It'd be very simple.

    11. DL

      Yeah.

    12. DV

      The trick is for this particular workflow, right, is, is substantiating it. As we zoom out, this is the case for, like, why do accountants and, and advisors and consultants exist, period, right? Because it's rarely about what math you can do. It's about what you can prove happened. And so for the R&D tax credit, for example, it's all about contemporaneous documentation to say, "Hey, how do you know Tom spent this time doing these things?" "Oh, hey, well, we pulled, you know, Git issues. We pulled Jira tickets. We pulled..." And then, "Hey, we can prove he spent hours doing this." Easy for a software company. Somewhat more nuanced perhaps for a manufacturing company or an architecture company or an engineering company even that might have different levels of tracking of these things.

  5. 4:18 – 6:26

    From RPA theater to frontier models: why automation works now

    1. TB

      Is this the reason this field is now able to be automated, whereas five years ago even it wasn't? Like, is it that the computers can understand words and read documents in a way that unlocks those types of customers?

    2. DV

      Yeah. I think a large part of it is that, right? So when I, when I was at Grant Thornton for, for all those years, I brought up to them and said, "Hey, you know, there's, there's technology that can achieve these outcomes." Obviously, I, I was not, uh, thinking of what we have today, right? I, I was a young guy. I said, "But hey, at its absolute worst, we can automate certain aspects of this workflow, right? Certain like repetitive tasks." This was in like the kind of like RPA quasi res- renaissance, right? This was, you know, some of the companies I knew spent like millions of dollars on like Alteryx licenses. I'm really just dumping on a lot of companies today. Point being is that they bought all these RPA soft- I remember like last year at, at Grant Thornton, I'm laughing, they had this thing called Automation Anywhere, and they wanted everyone to become a certified Automation Anywhere expert.

    3. DL

      [laughs]

    4. DV

      Which is just like the funniest thing in the world. And, and so you do this Automation Anywhere certificate, and like you just like play the video in the background, you get a certificate. And like the firm's like, "Guys, we did it." And it's like, "Yeah. Did we?"

    5. DL

      Yeah.

    6. DV

      Right? And so in the past, you know, Dave, you, you, you could have automated a lot of tasks. I think now what we've seen in the last, you know, three... I mean, our batch time, you remember our batch started in January '23, which was like six weeks after ChatGPT came out.

    7. TB

      Mm.

    8. DV

      Right? So at, at our company, at Onshore, I was, uh, just like fortuitously enrolled in like the GPT-3 beta in like early '21, and it was like, it was like pretty bad, right?

    9. DL

      Yeah.

    10. DV

      It couldn't achieve the outcomes today, but I could see at the time like, hey, if this gets just like a little bit better, this is just as good, if arguably not more capable than like your junior and even your mid-level accounting staff. And so now, you know, having arrived at where we are today-With these incredibly capable frontier models. I mean, the fact of the matter is that today is the worst they'll ever be. And so we're already at a place where I, I personally believe that they're outperforming people at the junior, the mid-level, and, and certainly kind of touching, if not exceeding, the capability of some of, like, your most senior technical experts. I mean, and, and, and that's kind of the reality of where we're at.

  6. 6:26 – 7:47

    Founder backstory and Onshore’s mission: automating tax, then audit and advisory

    1. DL

      Tell us how you got to, to Y Combinator in 2023. Like, what was the backstory?

    2. DV

      Somewhat non-traditionally, I would think. Uh, and so before I did YC in '23, I worked at a big, giant accounting firm called Grant Thornton, and I went to school for, for accounting and computer science here at, at the University of Illinois, so I'm from the Midwest, right? Went to school, you know, for arguably a non-tech thing. I mean, I have a s-

    3. DL

      Well, you had the intersection, right? You had-

    4. DV

      Yeah. I, I had-

    5. DL

      ... accounting and computer science.

    6. DV

      Exactly. I had the CS, uh, the CS background, and I had the accounting background, and then kind of inertia took over and said, "Hey, I, I got all these hours to be a CPA, so I may as well just, like, finish it," right? Graduated, went to go work at Grant Thornton, and were introduced to maybe the less conceptual and the more executional work, which was less exciting.

    7. DL

      And so tell us, what does Onshore do?

    8. DV

      So at its core, Onshore uses technology, AI models, uh, to automate a lot of the work that is being done today by accounting firms. And that's things like tax and, and hopefully soon things like audit and even advisory services, right? To take away a lot of the repetitive, superfluous tasks that people are paying a lot of money for, and automating that so you can really rely on the expertise, the capabilities, and genuinely, you know, get what you're paying for from an outcomes perspective.

    9. DL

      So you were back at Grant Thornton. I don't know how... Was it six, five, six, seven years ago?

    10. DV

      It was five, five... Uh, I was there, yeah, about five years ago.

    11. DL

      Five years ago.

    12. DV

      Yeah.

  7. 7:47 – 11:10

    Why incumbents resist: billable hours, partner incentives, and ‘AI spend’ as PR

    1. DL

      And at the time, you were kind of trying to suggest they use software, and it sounds like they weren't very receptive to it. Do you think today they are trying to transform their own businesses, trying to disrupt themselves?

    2. DV

      The idea is, do I think that they want to do this? No. And I don't think, I know they don't, right?

    3. DL

      Mm.

    4. DV

      So I have some friends, not at Grant, but that work at other firms. I actually talked with, with somebody at the largest of these big four firms. And they... There's all this press that's, oh, you know, Deloitte investing a billion dollars in AI, and, and, and KPMG and, and Cherry Becker at one of the top 20 firms. And it's a lot of noise. It's a lot of rattling of coins to say, "Hey, we're doing this." But when you get down to it, I, I learned at, uh, one of these large firms, the investment was purchasing, like, uh, three to four million dollars of Copilot licenses so that everyone could have Copilot. And so I said, "Well, okay, what do you guys do with that?" "Nothing."

    5. DL

      There you go.

    6. DV

      "It's horrible. It doesn't work."

    7. DL

      Yeah. I spoke to, um, a couple of guys who, uh, were the head of AI, joint head of AI at one of the, uh, big accounting firms.

    8. DV

      Sure.

    9. DL

      And they were trying to use AI to transform accounting and audit, uh, which by the way, I love as an idea. Um-

    10. DV

      Yeah. Concept, great.

    11. DL

      And I asked them, um, how many software engineers they had working with them.

    12. DV

      Oh, zero. I can tell you from my experience when I was, when I was at the firm, we had me for, like, three years. There, you know, of all the United States Grant Thornton employees, I was the only one with a software background, which is horrible. And then my last year, we tried to build something called, uh, tax innovation. And so we hired, like, three other people with a tech background, who then all subsequently got laid off during the pandemic. I bet if you were to ask any of them today, you're gonna get the exact same answer. "Oh, we don't, you know, we don't have any, but we are learning cloud code."

    13. DL

      Yeah.

    14. DV

      Or, "Hey, we, we, you know, the, the co-work thing that just came out." Or even worse, they'll say, "Yeah, we're actually paying a development agency, you know, somewhere overseas to, like, build this." I'm like, "Guys, this is never gonna work," which is what I experienced when I was there. I said, "We should try this." Said, "Oh, yeah, yeah, but, like, try how? Like, maybe can we build a prototype in Excel?" It's, "No, we can't." Like, but just believe that this could work. And fundamentally, I think, Tom, like, you know, using AI technologies to disrupt, you know, auditing, accounting, tax, whatever, is, is a great idea, but I think a lot of them like to say it, but the reality is that, like, they don't wanna do it-

    15. DL

      Mm

    16. DV

      ... for two reasons. One, it is a fundamental shift in what they perceptionally drive as value. Right now, they bill for hours of expertise, and as soon as, like, the hour becomes a diminished unit of value, you no longer really have a whole lot of ground to stand on when you say, "Hey, certainly pay me 1,000, 1,500, $2,000 an hour." And then moreover, the reality of the situation is that you have to then, as a senior partner, a guy, 'cause a lot of these firms have a mandatory retirement age. You gotta say, a guy who's maybe has three to five years left in his career, "Hey, we as a firm, that's gonna affect my pension, my payout for the rest of my life, have to plow a ton of money into something I will never reap the benefits from."

    17. DL

      Mm.

    18. DV

      And so you have this real asymmetric value prop. You got, you know, some 40-year-old partner says, "Hey, let's use AI." Some older guys, you know, at 63, 64 years old, say, "Well, I'm gonna retire in 24 months. I don't, don't care about this. Like, I'm not gonna see the value here in, you know, five, six, 10, 15 years." And so quickly the guys who make the decisions, the senior partnership, say, "Well, what if we just don't do that?"

    19. DL

      Mm.

    20. DV

      And then you get into this cycle of, of people like me who work there, who wanna make a change, who suggest these types of changes, and get shut down and say, "Okay, fine. I'll just leave and eat your lunch, and, and that's fine with me."

  8. 11:10 – 13:37

    The failed go-to-market: selling automation to accountants who fear fewer hours

    1. DL

      So you left. You-

    2. DV

      I did

    3. DL

      ... built this product.

    4. DV

      Yep.

    5. DL

      Um, you have a choice to try to sell it to those customers to automate their work-

    6. DV

      Yes, we did

    7. DL

      ... or go compete with them.

    8. DV

      Yep.

    9. DL

      Um, it seems like now you're doing the latter.

    10. DV

      Yes.

    11. DL

      But you started with the former. Walk us through that, like, journey.

    12. DV

      Yeah, yeah. So I, I left the firm in, in 2020 in August, and so my initial idea, right, was, well, I, I can fix the problem that, like, that I observed here. And so I built some software. I, I created something that I felt was pretty compelling, and then went to these firms, say, "Hey, listen, how'd you like to make this, you know, 80% faster?" We had, at the time, probably eight customers or so, like, doing some, like, mid six-figure amount of revenue by the time we, you know, cut over two years after this whole, uh, transition happened. But for those two years, the, "Hey, listen, you know, these are top 20, top 50 firms that wanted to implement our product." And when you pitch that, you know, Dave, to, like, a senior partner, they hear, "This is great. I'm gonna bill the exact same amount, but my margin is gonna go through the roof." Because when I was at the, the firms and, and my compatriots at these firms, if you have a project that's, like, 30% gross margin, like, you're a hero.

    13. DL

      Mm.

    14. DV

      Like, that's the apex of achievement.

    15. DL

      Mm.

    16. DV

      And when they think, "Well, hey, hold on a second-80% reduction in, in cost, margins going up. I'm the best partner at this firm. Problem is rarely are these senior guys making the decisions actually doing the work. And inside of all these firms, there's this really weird like asymmetric value proposition, which is you need to achieve a certain number of billable hours to stay employed at the firm. And so if you're like a junior guy, a mid-level guy, someone trying to become a partner, and I walk in with this offer, which I did for two years, and say, "Hey, listen, automate 80% of the work you're doing," they hear, "I'm gonna, I'm gonna get fired. I'm not doing this." And so for two years, I just... I fought with these guys. "Hey, tech is better. AI is better. This is more interesting." And I would get pushback constantly. "No, no, my pivot table is better. This is an edge case."

    17. DL

      Mm-hmm.

    18. DV

      "It doesn't fit. We do it this way. Change your tech to fit this mold." It's like, "Guys, just try it." And, and so after like two years of like banging my head against the wall, and this is how we ended up coming to YC, I was just like, "You know what? I don't wanna build that company anymore." Right? Selling software to accounting firms and consultancies is a losing game. And I, I'd say that to everybody. People ask me all the time, "Well, can I build for software for accounting?" Maybe. And like maybe there's a billion-dollar company out there to be made. I don't see it because your customer is truly the least motivated group of people on the planet to achieve what you're trying to achieve.

  9. 13:37 – 15:55

    Why legal AI adoption differs: perceived risk and shifting billing models

    1. DL

      This all makes sense, and it, it seems very similar to what we saw when companies tried to sell to lawyers previously.

    2. DV

      Yeah. Exact same thing.

    3. DL

      But that seems to have been solved, right? Lagora is on a tear. Harvey's doing really well. Um, they've figured out a way to actually sell to people who bill by the hour.

    4. DV

      Right.

    5. DL

      Do you know the reason why it worked there and not in accounting?

    6. DV

      There is still a perceived premium on like legal time. 'Cause like lawyers, you know, you mess up, massive fine, jail maybe.

    7. DL

      Okay.

    8. DV

      Right? Accountants, you mess up, like business inefficiencies. When it comes to tax, maybe jail, but like very rarely. And so I do think there's still an industry difference perception. I also would contend that a lot of these companies are on a tear because lawyers are maybe more accepting of like the incoming new world order, as opposed to accountants who perhaps have their heads like slightly more buried in the sand.

    9. DL

      Lawyers are forward-thinking. It's the first I've-

    10. DV

      Yeah. That, that might, that might be the hottest take of the whole thing.

    11. DL

      I think my observation of law is that, um, the industry has been moving to project-based billing for some time.

    12. DV

      Yeah.

    13. DL

      This idea that everything is billed by the hour is just not true anymore.

    14. DV

      Mm-hmm.

    15. DL

      And even when I was at my last company, we hired auditors or accountants-

    16. DV

      Yeah

    17. DL

      ... and they would bill by a project. You know, they'd say-

    18. DV

      Oh, yeah. Yeah, we did great. We did, we did a lot of project, like fixed fee project stuff.

    19. DL

      And I think basically law will go that way, accounting and audit will have to go that way, and once you're, you're, you're billing by project, you have price competition, and then people have a lot more incentive to, to reduce their cost.

    20. DV

      You would think.

    21. DL

      [laughs]

    22. DV

      You would think, because... And I say that because a lot of the stuff that we did, and even a lot of the stuff that like when I was selling to accountants was project billed, right?

    23. DL

      Yeah.

    24. DV

      Even our... I mean, ours is like a whole other issue. But project billed stuff, they still don't wanna reduce those fees. And so when they come into these pitch, you know, "Hey, we might be 20% more expensive, but we have the smartest people doing the smartest work." And, and I always say this, that is probably true for one firm, right? Like only one can have the smartest people. So let's say it's Deloitte, for example. By definition of smartest, like everybody that not there can't also have the smartest people. And so, you know, you get this really weird dichotomy where it's like, "Hey, our prices are, you know, project-based." And I'll bet you if you divide the project by the number of hours that it takes, that it gets real similar to their hourly rate.

    25. DL

      For sure. And that's how they come up with the project fee-

    26. DV

      Sure

    27. DL

      ... to start with, I think. So you tried selling directly to these tax advisors, these accountants, and you banged your head against the wall for two years.

    28. DV

      Yeah.

  10. 15:55 – 17:45

    The pivot: firing firm customers and selling directly to corporations (end users)

    1. DL

      And then something changed.

    2. DV

      Yeah. Basically, after two years, I got so frustrated with these people, like, 'cause I, I knew, right? I saw it. I believed it. I lived the problem.

    3. DL

      Yeah.

    4. DV

      I did the thing. So I knew what was possible, and I built the thing that was possible. And to have a lot of people just tell me directly, "This isn't right. This doesn't work. This will never happen," was just like infuriating because I, I, I don't know why, but just at like the core of my being, I just knew this was gonna happen. And I knew that, you know, why can't I be the precipitator of this, of this thing? And so after being told this so many times, I thought, "Well, maybe the accountants aren't actually the ones that derive any value out of this," right? Like when you just zoom out, like who actually gets the benefit of, of doing an R&D tax credit or a, a tax return even, or an audit, right?

    5. DL

      Yeah.

    6. DV

      The company. The government kind of regulates a lot of it. There's rules and regulations and laws, and so really it's just a, an transaction, a contract between the government and the company. And accountants have kind of just wormed their way into the middle and artificially sel- set themselves up to be, you know, "I'm the middleman. I'm the expert." I'm like, what if we just didn't have to have that?

    7. DL

      Mm-hmm.

    8. DV

      Right? What if we stopped enabling this artificial middleman and just like, you know, this whole like sell shovels in a gold rush thing, what if you just like burned all the shovels, broke them in half, and just brought the gold by the handful to the company? Like that seems a lot better to me. And so in summer of '22, we fired all of our customers, said, "Hey, sorry, we no longer offer the service. Like your contract's terminated. Thanks. Have a good day. Goodbye." I took my ball and went home, rebuilt the software to be focused on the corporations, the taxpayers, right? The people that ultimately benefit from this. And then started just going on LinkedIn and like cold DM'ing a bunch of people. And I mean, this was like old school stuff. We didn't have like all the cool AI sales, like silliness that we have today. So I was just like, "Hey, my name's Dominic. I can do your R&D tax credit better because I use AI." And I sent a message very similar to that to like a thousand people.

    9. DL

      Mm.

  11. 17:45 – 18:46

    Early proof and YC entry: the Miami deal and rapid acceleration

    1. DV

      We got this company in Miami to say, "Hey, you know, come down. You know, we wanna meet you in person because we don't know about this tech thing, but you seem like a genuine guy." So we flew to Miami. You know, I, we, I, we had no money. Uh, you know, stayed in some random hotel. They said, "Hey, this sounds really good. You sound better and more efficient than our current provider. It's a lower price. Let's give it a shot." And I said, "Oh, okay. Great." So they signed this like paper contract, like in person-

    2. DL

      Wow

    3. DV

      ... that I like put in a folder and like took home. And then like the... While we were there, like on the flight there, we got an email from, from YC said, "Hey, you wanna do this interview?" And we had applied to YC on kind of a whim. I knew about it from college, like from like my computer background. Had some friends that like had done it, but yeah, didn't really know much about it, so we applied, did the interview there, and then like before I got on the flight back home to Chicago, I, uh, we had gotten in, which was like-Arguably one of the most transformative things that ever happened.

    4. DL

      One of the problems sometimes is that when you are a big company and you are paying for a big logo like Grant Thornton or Deloitte or whatever-

    5. DV

      Mm-hmm

    6. DL

      ... there's some, there's this sort of, you know, no one ever gets fired for buying IBM-

    7. DV

      Sure

  12. 18:46 – 21:44

    Co-founder credibility and loss: building with a senior partner from Grant Thornton

    1. DL

      ... kind of thought process. Whereas, like you turn up and, I mean, your co-founder... Why don't you tell us a little bit about, about Mark, your co-founder?

    2. DV

      So I, I did have a co-founder named Mark. Um, he was actually my like boss's boss's boss at Grant Thornton. And we had worked together on a few projects, um, had collaborated a bunch and, and he was a, he was a reasonably forward-thinking guy for like a senior partner at this accounting firm, right? He, I believe, really saw what I was pitching there and was confined by his circumstance. And so when I left to start this company, I had called him up, you know, uh, probably a few months, maybe a year later and said, "Hey, listen, I'm getting some traction. I'd love for you to come over and lend, A, some professional credence to this." You know, senior partner at Grant Thornton comes over and co-signs this innovation, becomes somewhat more palatable than like junior guy, uh, you know, that started doing this. But then, B, just a wealth of, of expertise and knowledge. And so when he came over, we worked together for a few years. Uh, he contributed quite a bit when it comes to the tax knowledge, you know, how we actually approach these companies, you know, from a value perspective. Uh, and then unfortunately, he, he passed away in the summer of 2024, uh, about a year and a half ago, uh, kind of right on the precipice as we were raising our series A, so.

    3. DL

      Yeah, that's very sad.

    4. DV

      Yeah, it was sad. He was, he was a great guy and, and really enjoyed working with him. You know, I, I can't say enough how, how instrumental he was in the early days and, frankly, instrumental he was in my career at Grant Thornton, right? I learned a to- a tremendous amount from him and, and appreciate everything he's done for us.

    5. DL

      Yeah. So, so Mark was like the senior partner, kind of-

    6. DV

      Yeah

    7. DL

      ... had he, he'd already retired?

    8. DV

      He was like right on the edge of retirement.

    9. DL

      Right. It's kind of an unusual team configuration. We don't-

    10. DV

      Yeah, very

    11. DL

      ... we don't see many YC companies with a, a senior partner from Grant Thornton and like a entry-level associate. But with you guys it seemed to work.

    12. DV

      I would love to tell people like there's a magic, like bullet or sauce or something, secret sauce behind that. The reality is that I think I had a pretty good idea, and I was a reasonably convincing guy. And so I just called him up. We probably had three or four phone calls about it, on the phone for an hour or two, and I kinda pitched him, "Hey, here's what I wanna do at the beginning. And then if we nail this, here's what we can do next. And then if we nail that, like then three or four steps down the road, like there really is this kind of new world order where like accounting firms kinda can cease to exist, and like it can be only taken, you know, by Onshore as technology and then, then models and things like that."

    13. DL

      I mean, obviously, like Lagora has done this without hiring a senior partner from-

    14. DV

      Yeah

    15. DL

      ... a law firm. How instrumental do you think that credibility was? Do you think you could have put it any other way?

    16. DV

      Yeah. I, I think it's possible certainly that I think I coulda gotten it done. Like you said, Lagora's a great example. I think it would've been possible. I think it would've been appreciably more difficult.

    17. DL

      Yeah.

    18. DV

      Right? A, and I also think that, uh, a lot of the buy-in from early customers, early taxpayers, was that we could functionally replicate with technology what they were getting before. And the reason we could do that is because I knew exactly what they were getting before, 'cause I did the work and Mark had sold the work, right? And so for us, we kind of knew all of the pain points, the levers, from just having lived in it so intimately. And so I do believe it's possible. I believe it would've been possible for our company. It would've been more difficult, and we probably would be less successful.

  13. 21:44 – 25:34

    Forecasting the new accounting pyramid: fewer juniors, more leverage, different roles

    1. DL

      Yeah. Let, let's play it out a little bit.

    2. DV

      Sure.

    3. DL

      Y- your company's doing really well.

    4. DV

      Yeah.

    5. DL

      Growing really fast. If you continue this growth, what impact is it gonna have on the field of accounting? Like, a- and the, your previous employer, the other companies, are they just gonna like slowly recede to nothing, and they will be gone? Like, what's your view on, on how this plays out?

    6. DV

      Yeah. Okay, from a company perspective, I do wanna, I w- I wanna split this out because I don't wanna seem like an anti-worker, like I'm trying to take a million jobs from people. But the reality is that like that's the point, right? That's the hope. You know, some of these firms have existed for almost as long as like America has existed. Like, I believe Deloitte was founded almost 250 years ago. From my perspective, I understand why it's hard to imagine a post-big four or post-accounting firm world. It is a foreign concept. Conversely, I don't actually think they've earned the right to maintain the, the reverence that they've, they've been granted for all of these decades, centuries sometimes. And so if this keeps going the way that I think it's gonna keep going, the way that I'm kinda betting the house that it's gonna keep going, is that's exactly what we're looking at, right? A, a functional and tectonic shift in the way that these outcomes are ultimately provided to customers, because that's, that's all we're doing. That's all we do. That's all they do, is we provide an outcome, right? They provide it, you know, using billable hours and projects and manual processes and humans and all these things, and we provide it slightly differently.

    7. DL

      Do you think the accounting industry will make more or less revenue in 10 years?

    8. DV

      I think the industry will make more revenue.

    9. DL

      And do you think it will employ more people or fewer people?

    10. DV

      I think that it will probably employ a similar number of people. And so what I don't necessarily see happening is, is a complete elimination of the need for humans in this cycle. And so the reason I say that, and, and I'm not even saying the humans that currently exist in the cycle, 'cause I think a lot of folks that are currently in it won't be in it in 10 years, right? I do think that this will enable a new type of professional, a new type of human professional, to enter this. People with a more technical background, right? People with a, a higher kind of threshold of innovative thinking, right? To enter this and, one, manipulate some of these tools, work in concert with some of these tools, and then, two, and arguably more interestingly, really provide what these firms were set up to do to begin with, right? Because when, you know, these firms were founded many, many decades and centuries ago, they really were a, a haven for the experts in, in specificity to explain that to somebody else or to achieve an outcome for somebody else.

    11. DL

      If the metric is revenue per employee-

    12. DV

      Yes

    13. DL

      ... I imagine that is gonna go through the roof.

    14. DV

      Yes.

    15. DL

      And I don't think total spend on accounting is gonna 10X.

    16. DV

      Right.

    17. DL

      And so therefore, the thing that's gonna flex down is the number of people.

    18. DV

      Yeah, therefore it's gotta be headcount. Yeah. I think revenue per employee at Deloitte will go up, but I do think that companies, taxpayers, corporations, will have more people that do more of these things internally.

    19. DL

      The accounting firm of the future-

    20. DV

      Yeah

    21. DL

      ... first of all-

    22. DV

      Has less people

    23. DL

      ... well, first of all, it will change shapeSo if you look at Deloitte or KPMG or... It looks like that, right? You have a-

    24. DV

      Yeah

    25. DL

      ... a small number of senior partners. Yeah, lots and lots and lots of juniors.

    26. DV

      Yeah.

    27. DL

      I think it will end up more like that.

    28. DV

      Yeah.

    29. DL

      Um, I don't think you'll have the very fat bottom layer 'cause I think you'll have AI agents doing it.

    30. DV

      Yes, sir.

  14. 25:34 – 28:41

    Onshore’s scale economics and wedge strategy: start narrow, expand via trust

    1. DL

      ... today. Just people are growing to much higher revenue with far fewer employees faster. Okay, so tell us, um, you've grown revenue pretty quickly so far. Where are you up to right now?

    2. DV

      Uh, so this year in '26, we're targeting getting to 100 million.

    3. DL

      And you're w- today, you're at 20-ish?

    4. DV

      Uh, we're, like, 25 or so.

    5. DL

      25. Okay, so you went from 25 to 100 million in re- revenue this year.

    6. DV

      Yeah.

    7. DL

      And how many employees will you have at that stage?

    8. DV

      I think by the end of the year, we'll have, you know, somewhere between 60 and 100, probably close to 75.

    9. DL

      Okay. Well, for sake of mathma- let's make it easy. Let's call it 100.

    10. DV

      Sure.

    11. DL

      So you're making a... You... By the end of the year, if you hit your targets, you'll make a million dollars of revenue per employee.

    12. DV

      At least.

    13. DL

      And so what is that number for someone like Grant Thornton or Deloitte or KPMG?

    14. DV

      My estimation would be probably 100,000, maybe 150,000.

    15. DL

      You're an order of magnitude better-

    16. DV

      Yeah

    17. DL

      ... and you will only get better going forward-

    18. DV

      Yeah

    19. DL

      ... because you're software-based, whereas they can't really change.

    20. DV

      Right. And they... I mean, arguably, they will only get worse, especially now that we see instead of investing into technology, we see a lot of these big companies investing into overseas operations. And-

    21. DL

      Mm

    22. DV

      ... and so that's just gonna dilute the revenue per employee even further.

    23. DL

      So you started in, like, a very specific thing, R&D tax credits.

    24. DV

      You got it.

    25. DL

      And I'm sure you have a plan to expand beyond that to-

    26. DV

      Yeah

    27. DL

      ... more general accounting. Do you think that's the right approach now that you've done it, like picking some small wedge and expanding? Or as technology gets more capable, can you just kinda, like, go and do it all straight away?

    28. DV

      I still maintain the way that we did it was the best way. Um, and I do think that, like, the, the ever-increasing capability of technology makes it easier to do it the other way. I kinda think that's a bit of a fool's pursuit, right? The idea of trying to boil the ocean all at once is very challenging, right? And when I look at the way that we work with customers, the way that people interact with us, and ultimately purchase products from us, it has been an incredible benefit for our business, for Onshore's business, to be great at one thing-

    29. DL

      Mm-hmm

    30. DV

      ... really early. And, and, like, obviously we've expanded now, and we're gonna expand in a tremendous way over the next, you know, many years. But the reality is when we can come and show, hey, we can do an R&D tax credit better, faster, more affordably with better traceability, if we come up and say that and prove it, right, it's way better than showing up and saying, "Hey, we can do all your stuff," right? Because rarely do you have a CFO at a, at a mid-market or enterprise company say, "Hey, you can do all our stuff. All right. Somebody scrape out Deloitte. These guys are starting it," right? But if you show up and say, "Hey, we can do your R&D credit a little bit better." "Well, hey, we'll give that a shot." And then they experience it, and the next year it's, "Well, hey, you did great on this R&D credit." And I'll tell you, the way that we have built all of our new features, all of our new products, has been this interaction. "You did this well. And because of that, can you do that or this, X or Y or Z?"

  15. 28:41 – 31:29

    Beyond Onshore: the next disruption target might be Excel for knowledge work

    1. DL

      If you were gonna start a startup today, what idea would you choose to tackle?

    2. DV

      I'd love to tell you I have one specific thing that's gnawing at me, but, but the overarching thing that keeps, I keep coming back to is a better or different, uh, Microsoft Excel, right?

    3. DL

      Hmm.

    4. DV

      Uh, and the reason for that is I think spreadsheets and, and Excel in its original iteration and in many iterations down the road as, as a spreadsheet functionality is great. I think that it has been kind of just, like, squished and torn and fit into a million different boxes for things it was never meant to do, and we have a, a huge part of our professional society, both in, in accounting and legal, banking, built on this tool that really was never meant or intended to be built upon in the way that it has been, right? And so I think I, I don't have a great answer to what the software should look like or what the features should be, but I do think a, a reimagining of, of the way that some of these industries work on Excel and, and a supplet- supplantation or, or new, net new, like, workflow there would be really interesting to see.

    5. DL

      There are obviously vibe coding tools. You could go and use Replit, but why aren't your customers using something like Replit?

    6. DV

      Yeah. I mean, even to some of the stuff that we talked about earlier, right? I think if you, you ask somebody, "Hey, why don't you vibe code a Replit app?" They would ask, "What is Replit, and what is vibe coding?" Uh, and so I think that the, the access and the democratization of that knowledge is less permeating society than we all might think it is. But the other side of that is that, like, a- as we talked about, like, the, the agency piece of it, right? If, if I was, you know, kind of a junior guy at Grant Thornton, and I wasn't me, and I was just some of the folks that I had worked with and said, "Hey, why don't you just vibe code this app, uh, or use Replit to build this app to, like, fix, like, the billing system at one of these accounting firms?" My immediate answer would be like, "Yeah, I'm not gonna do that."Right? That seems like a lot of work, and I'm not actually gonna get anything for it. They're just gonna take my thing, and I'm just gonna like go play basketball on the weekends. And so I think there's capability there in the existing tools. I think that the knowledge gap, the agency gap exists. And also siloed and specific apps for everything that Excel is built on leads to this like bloat of, of functionality or a bloat of individual tools-

    7. DL

      Mm

    8. DV

      ... which like is a problem in and of itself, and then some other guy that comes on this podcast a decade from now will wanna fix that problem.

    9. DL

      Okay, so you think there's space somewhere between Excel and Replit-

    10. DV

      Yes

    11. DL

      ... for like the non-technical professional services user to take, to like rip all the functionality out of Excel and b- and build it in a way that they can kind of understand and maintain.

    12. DV

      Yes. That, and also not rip out individual functions, but rip out macro-type functions. And macro in that, you know, you could rip out a, a subsection of five, six, seven, 10 things that you're doing in Excel that may be seemingly unrelated, but ultimately use the same foundation of technology, whether that's a pivot table, whether that's, you know, VBA or, or, or databasing, right? And bring that into its own tool, right? So you don't have to have 50 apps. You could have five.

  16. 31:29 – 33:55

    Adoption curve: capability is here; acceptance is catching up; jobs shift later

    1. DL

      So this has been a really interesting conversation. L- Last year we did a couple of shows where we talked about the impact of AI on software engineering and, and made some predictions about what would be possible, and I, it seems like those predictions were not a million miles away what it, th- compared to what's actually happened.

    2. DV

      Correct. Yeah. In terms of the technical capability of these AI systems, I think w- we were kinda right about that. We have not seen mass layoffs in the software engineering world.

    3. DL

      Yeah. That, and that's interesting. So like the models have continued to get good. The capabilities match what we thought they would sort of do, and, and even in the last few weeks, you can kind of sense a bit of a shift, you know, if you spend all of your time on, on Twitter like I do. Um, great professional software engineers are now starting to embrace AI, right? And like I think the contention that software programmers will no longer write code sounded crazy nine months ago, and now is like orthodoxy almost.

    4. DV

      Yes.

    5. DL

      And my contention is the same thing is gonna happen to accounting and law and audit.

    6. DV

      Yeah.

    7. DL

      The things that we thought you needed humans for, we won't need humans for, uh, for very much longer. And so by the end of this year, I can see, um, mass realization that the AI can do these knowledge work jobs just as well or better than humans, and yet still not that big a change in employment rate.

    8. DV

      Right.

    9. DL

      I think that might be like five or six or seven years to come.

    10. DV

      No, I, I would agree with that. I think... I mean, I, I'll be honest, I mean, even, even from our internal, right? This year, you know, second half of last year even, we see a lot more excitement and acceptance of, of the AI portion of our business, right? In the early days, you know, "Hey, we have AI, but don't, don't worry about that. It's just doing some..." And now people are excited about it.

    11. DL

      Mm.

    12. DV

      Like, people, people wanna see it. We had a few, a few customers saying, "Well, hey, can I see your AI do the work?" I'm like, well, that demonstrates that you don't exactly know how AI works, but like that's fine. Like, I, I appreciate your, your ambition, right?

    13. DL

      Enthusiasm.

    14. DV

      Enthusiasm, exactly, and that's what we kind of predicated on. And so I think that, I think you're spot on, right? I think the end of this year we'll see not just realization, but a lot more like fundamental acceptance that AI can achieve these outcomes either at the same level or, or arguably better than they have.

    15. DL

      Mm.

    16. DV

      I don't think that's gonna lead to like mass unemployment in 2026 or even '27.

    17. DL

      I agree with you. Look, this has been a fantastic discussion. Thank you for joining us, Dominic. It's been a pleasure.

    18. DV

      I appreciate it, guys. Thanks so much. [upbeat music]

Episode duration: 33:55

Install uListen for AI-powered chat & search across the full episode — Get Full Transcript

Transcript of episode lfzm2SlhbM8

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

iOSAndroidClaudeChrome