Basis's Matt Harpe: Why AI Won't Replace Accountants — It'll Make Them Richer

0:00Matt Harpe:I do think that accounting is extremely important. It's underappreciated. And I think that this North Star of, hey, we're going to help, you know, organizations like, you know, health systems that I used to work with truly understand how the organization functions and allow them to measure the core goals that they should be tracking against so that they can make better decisions. To me, I do think that's a very important part of how our economy works. The U.S. has 3 million accountants, and they all have a problem. They're swamped by busy work. An AI startup called Basis wants to help those accountants spend more time on the stuff they really care about, advising clients, getting more work done, and keeping all of us in the good graces of the IRS.

0:36Toward that goal, Basis recently raised$100 million at a$1.15 billion valuation. And our guest today is the co-founder and CEO of Basis, Matt Harp. He says that as the number of accountants shrinks, he can help firms do more with an assist from AI.

0:50Matt Harpe:When we started Basis, I think very few companies and certainly very few accounting firms were interested in doing AI for the sake of doing AI. Now everyone's board, every leadership team is thinking about how do we incorporate AI into what we're doing. But that wasn't true when we started. Today on the podcast, we're going to talk about why accounting is the perfect vertical for AI impact, how to recruit top talent in an unsexy problem like accounting, and why Matt thinks he can win where a whole wave of startups didn't succeed. plus how basis can compete as Claude and ChatGPT inevitably move in on its clients.

1:24I'm Alex Conrad, founder and editor of Upstarts Media, and this is the Upstarts Podcast, our weekly show where we talk to startup founders who punch above their weight to take on the status quo. Matt, welcome to the podcast. Alex, thank you for having me. It's a pleasure to be here. This episode is brought to you by Rippling AI, the only AI built to give you full visibility into your startup and the ability to take action across every department. So what is the basis of Basis?

1:50Matt Harpe:Great question. So we like to think about Basis as the most accurate and capable AI for accounting. We serve accounting teams across the country. We're currently in about 30 % of the top 25 accounting firms and about 25 % of the top 150 firms in the country. And we're growing quickly. So you started this company with two co-founders in 2023. Was accounting something you were always deeply passionate about? Or where does the origin of Basis come from? Yeah, so Mitch, I started the company with Mitch in 23. Mitch has some background in accounting. He studied accounting undergrad and was at an accounting technology company.

2:22Matt Harpe:So he had a fair amount of background there. I have no background in accounting at all. I got introduced in a relatively circuitous manner. My interest sort of academically was on machine learning topics in healthcare, was focused on a set of mortality trends around the opiate epidemic in the US, ended up doing a whole bunch of work there while I was in undergrad, and then went to the Boston Consulting Group to sort of work on sort of similar methodological type problems in and around healthcare, spent a lot of time working with health systems and pharma companies. And ironically, through that process, I ended up getting introduced to accounting.

2:55Matt Harpe:So we oftentimes have to interface with accounting teams. And this actually had nothing to do with what we were working on at the time, but I had a kind of crazy realization in that process, which is that I kind of learned that if you go to any hospital in the US and just ask what I thought was a very simple question, which is how much does it cost us as a hospital system to provide a procedure to a patient? There's really no hospital in America or very few hospitals in America that can actually answer this question. And this is just ridiculous because if you're a hospital, you should have at least two goals in my mind.

3:24Matt Harpe:Increase the quality of care for your patients and reduce the cost of care for your patients. And if you don't know how much things cost in the first place, you really have no idea if you're actually improving or getting better or getting worse at one of your core outcomes over any given period of time. And it turns out that, at least in our view, this is largely an accounting challenge. The job of accounting is to structure and systematize all the economic activity that happens in and around your organization so that you can understand that information, you can do things with what's going on.

3:51Matt Harpe:But there's only so much time in the day and accountants only have so much capacity. And so the 3 million accountants that we have in the U.S. predominantly spend their time working on accounting in and around what we call the compliance activities. You have to do certain structuring and systematizing to prepare a set of financials to be audited so people can trust what it is that you're reporting to prepare a tax return or a 990 or some kind of governmental filings that you can participate in our system of taxation. But there's all this other accounting work, some of which gets done, but a lot of which doesn't, which involves other forms of structuring and systematizing this economic activity that's actually quite valuable, that can allow you to better understand your organization and different aspects of it and how it may be changing over time and then help you in making decisions about your organization that just doesn't get done.

4:39Matt Harpe:If we had 30 million accountants, maybe we could go do all that work, but we have 3 million and actually the number is rapidly declining. We have a pretty significant shortage of accountants even to do the work that exists today. And so I became quite interested in this topic in terms of the sort of foundational aspect of accounting and how most organizations operate. And ultimately, the way that our economy is organized is that we sort of decentralize decision making down to these various organizations that are then sort of taking local inputs and making decisions based off of them. And accounting is really the language we use to sort of structure and systematize those local inputs.

5:12Matt Harpe:And so I became quite interested in accounting from that aspect. And then through conversations with Mitch, started to learn sort of the state of accounting technology and the extent to which there were many processes, which even at that point felt like some of the emerging abilities of these large models could start to help with. And we felt like it would just become clearer and clearer over time. And so we felt like there was a big opportunity to make accounting a lot more efficient and effective in what it did, but also just to sort of help increase the total volume of accounting in the world.

5:41Matt Harpe:And so that was kind of how I first got introduced to accounting and we started working on some things. And then one thing led to the next. Why should people be rooting for accounting AI to succeed? Why is your product needed so badly in this wider ecosystem, you know, really zooming out? The first way to look at it is just if you look at the accounting work that absolutely has to get done and that historically has gotten done, And core things like companies being able to produce financials, those financials being able to be audited and verified. Our whole system of taxation runs off of accounting.

6:14Matt Harpe:And a number of the ways that we understand all the various organizations in society are all downstream of accounting. And so a lot of that work is actually at risk because we have a relatively significant shortage of accountants in the U.S. There was a period of two years where roughly 300 ,000 accountants retired, which is a pretty significant number of people. the number of people studying accounting is not going up. In most years, it's gone down. I think there was actually a little bit of a blip and it went up for one year, which was fantastic. Demographically speaking, they say you never know exactly how to interpret these demographic estimates, but maybe 75 % of all accountants will retire in the next 10 years.

6:50Matt Harpe:And there are not enough folks to replace them. And so really the sort of core way that a lot of our economic infrastructure, that our economy and our broader society are organized is at risk because we don't have enough accountants. And so maybe, you know, challenge number one is, is there a way to help offset the declining population of accountants by allowing AI to help take on a bunch of those workflows? Why are there no accountants? Is it because we've offshored a lot of this work, or it's just not considered, you know, high salary valuable enough? Or is it just miserable work? Like, why is there such a gap to the best of the research you've been able to find?

7:25Matt Harpe:Yeah, so I think, again, there are multiple factors that converge there. Some of these are relatively longstanding trends in terms of the things that people, the different careers that people have been going into. But I do think one thing that probably should not be under underrated is, you know, being an accountant is a tough job. And the ups and downs, the sort of seasonality of the profession, you know, lead a lot of folks to get frustrated. And I think to consider moving on to other professions or to retire early, I think there were some really interesting stats that came out of the sort of the recent tax season.

7:58Matt Harpe:Tax is probably the most seasonal of all the accounting disciplines because obviously it's concentrated around certain windows of the year. And the number of accountants who experienced either burnout, mental health challenges, sort of having thoughts of leaving the profession is remarkably high. I think it's over 50 % of said surveyed accountants. I don't know exactly if that number is right or not, but we can say that there are a lot of people who feel this way. And I think a large part of the reason that's true is because of the seasonality, but also because so much of the work that accountants do.

8:28Matt Harpe:I think accountants feel like, how is there not technology that can help automate a lot of the more manual aspects of these things, the more grueling aspects of the work? And you probably would think that in the year 2026, we would have technology that can do this. I think Basis is solving a lot of those problems and is really the first technology to actually solve a lot of these things that you'd expect to be automated. But historically, the technology hasn't been there to do it. And I think as we're able to show accountants that they can spend time on the parts of the job that they really enjoy, advising clients, thinking about strategy, orchestrating these systems that are working to produce a great result for their clients or for their company, more and more people will actually want to stay in the profession.

9:06Matt Harpe:And hopefully we can convince more and more people to join the profession. But I think historically that has been a big source of why people have been frustrated with their experience over time and why a lot of people are considering or have left the profession. now my experience with accountants is that they are overwhelmed don't have time to necessarily play around with a lot of tools they're looking for very basic things like how can i share files with them a little faster but but they're not necessarily these accounts that upstars has worked with they are really busy people and they're maybe a little skeptical of you know whatever new technological trend might come pitch them what has been the door opener or sort of the really aha moment help that Basis provides that gets customers excited?

9:52Matt Harpe:It's interesting because when we started Basis, I think very few companies and certainly very few accounting firms were interested in doing AI for the sake of doing AI. Now everyone's board, every leadership team is thinking about how do we incorporate AI into what we're doing, but that wasn't true when we started. And I think one of the reasons that we were able to so quickly build trust with firms, deploy to firms, and start to see some of the traction that allowed us to continue to grow the company was the fact that this shortage of accountants was really impacting a lot of the firms that we were working with.

10:27Matt Harpe:So if you are an accounting firm, the way that you continue to grow your business is you hire great new accountants, they come to join the firm, and now you're able to serve more people and now your firm can grow. And if all of a sudden you're finding it really difficult to hire folks and you have relatively high attrition rates, especially coming out of COVID and in the years sort of succeeding it, that becomes one of the top priorities for you to solve in the organization. And you don't have that many options for how to solve it. It's very difficult for you as one firm to go convince a bunch of people to start studying accounting and go through the process of becoming a CPA.

11:01Matt Harpe:So that is maybe not an option that's really on the table if you want to have immediate impact. You could potentially offshore these jobs to another country. And there's a whole set of complexity that comes with that. And that is a pretty big industry, right? I mean, there is a lot of offshoring. A hundred percent. And, you know, especially the largest firms have invested a lot in offshoring overseas. And that's been successful in some ways and unsuccessful in other ways, but certainly a huge component of the accounting industry today, even though we still have three million plus accountants in the US, more than one in a hundred American adults is an accountant.

11:31Matt Harpe:You know, because this was an issue that was, you know, top one, two, three issue for most accounting firm leaders. And we were able to say, hey, we think we can, A, help make a lot of these manual processes more efficient, which is going to allow you to continue to grow your firm despite the fact that you have massive hiring challenges. B, take away a lot of the sort of more manual aspects of the job that people don't enjoy that causes folks to consider leaving the profession or consider going elsewhere. And C, give you an opportunity to invest in your people so that they can actually learn the sort of most cutting edge technology, and then hopefully that is a competitive advantage from a hiring perspective, I think that got people really excited.

12:11Matt Harpe:And when we were able to show that we could actually make an sort of an aggregate impact on the entire practice, change the sort of efficiency profile of the entire practice, not just pick one sort of niche workflow that wouldn't have any total impact on efficiency, I think that made a lot of firms open to considering something that maybe wouldn't have been a top priority for them at the moment. We've had people on the show in all sorts of areas of software and also these sort of vertical softwares like, you know, Harvey in legal for people who are not accountants themselves. Can you give an example of the kind of annoying process or task that would have been really tedious to do manually and now is automated through basis in a way that, you know, makes them a lot more productive and happier?

12:54Matt Harpe:We often think about a lot of accounting processes or you have to take information that's happening, all this complex information that's happening in the real world. I work at an organization. I sweat my credit card for something or I sign a contract with someone. And we need to figure out how do we actually recognize that event in a way that we can structure and everyone can understand. And then once we have that information stored in a way that we can understand, in order to know that we've done things correctly, we need to go back and verify that that information was properly structured and systematized.

13:25Matt Harpe:So you can imagine that, you know, we might have thousands of these events and we now need to go back through all of them to make sure that each one of them is supported by, you know, a piece of corroborating evidence. And that can involve just pulling up documents side by side and looking at all those documents. And then when something doesn't make sense, going into a system and pulling up some other information and clicking around in a bunch of different in a bunch of these different stores of information. That is a very grueling process that involves, you know, you sort of double checking numbers and clicking a whole bunch of buttons.

13:54Matt Harpe:And basis can essentially just end and do that entire process, present the results to an accountant, show them what it's certain and uncertain about and allow them to very quickly triage their time to only the most important things. and then be able to sign off and say this thing is done. And so there are these types of processes that might otherwise take 10 hours, 100 hours, depending on the volume, that now can be done in a matter of minutes with extremely high levels of accuracy and still giving the accountants the information they need to know in order to be really confident in the things that they're doing.

14:24Was your goal to be an entrepreneur and you just sort of back into accounting because it is this big need in the healthcare space? Or what was sort of the mission that a younger Matt had for their career?

14:37Matt Harpe:I think the thing that I was always very interested in was one, building things. I think there's something very cool about figuring out how to create something that doesn't yet exist in the world and something that other people value. And that maybe means that there's some kind of counterfactual impact where like, if this hadn't been done, then maybe things would have gone in a slightly different direction. And so I think I've always just found that to be really exciting, whether it's building Legos or building some piece of software or building, I don't know, a book of Sudokos. There are a lot of things that I just think creating is fun and interesting.

15:13Matt Harpe:And so I think being part of some kind of organization that's building and creating things is always interesting. I think second of all, I think I quickly realized as I worked in various companies that being in some kind of organization where the inputs that you put in could directly correlate with the outputs that you got out, that you could sort of embark on a really hard challenge and you could sort of figure out how to apply maximum effort to then lead to something that was really exciting was something that I was interested in as well. And then I think finally, picking things that were hard and figuring out how to just get over the hump of something that's really difficult was also something that I always was drawn to in some respect.

15:55Matt Harpe:And so young Matt could have imagined going into a whole range of different types of roles or jobs. But I think it needed to have some combination of those types of things. It could have been something in politics, potentially. It could have been maybe even something in academia. But I ended up feeling like working at a startup was like the best combination of all these things. Very exciting, very fulfilling. And ultimately, as this sort of idea took shape, I thought, let's try this and see how it goes. But I do think there are actually probably a range of exciting ways to experience those different components that would have been pretty fun as well.

16:30With Upstarts, the nice thing about my job is I get to talk to different founders every single day. And if I go too deep on one topic, I can jump to another one, you know, from nuclear energy to AI to dogs. Was there any moment for you where you were like, do I want to spend 10 years of my life focused on accounting specifically? Like this is the area that I can stay passionate about. Or was it more just the problem was so big that you were like, it doesn't matter if it's accounting or not?

16:57Matt Harpe:Look, I think it's really important when you're building a company that the things that you're working on are very exciting to you. And you can wake up every day ready to tackle them. And you can also convince other people that this is something they should be spending your time on. I think there's probably nothing that's more important in building a company than being able to attract amazing folks and build a great environment around those people so that they can do their very best work. and especially at a moment where there are so many great things to be spending your time on, explaining to people why this is the thing that they should go spend the next five, 10 years of their career on is extremely important.

17:33Matt Harpe:And if you don't believe that yourself, it's very difficult to go do that. So I think that's absolutely key. And yes, I definitely thought about that to make sure that that was something I felt for what we were building at basis. I mean, to me, I think there are two components. The first is I do think that accounting is extremely important. It's underappreciated. And I think that this North Star of, hey, we're going to help organizations like health systems that I used to work with truly understand how the organization functions and allow them to measure the core goals that they should be tracking against so that they can make better decisions.

18:04Matt Harpe:To me, I do think that's a very important part of how our economy works. I do think that's incredibly important. And I think most, I guess, accounting technology companies throughout history, I think, have not been able to tell that story in a way that's compelling. But I think the work that is done there is incredibly valuable and underrated. I think the second thing is getting to wake up every day and work with these amazing models, these non-biological intelligences that I think you could argue are probably one of the most impressive and impactful things humans have ever invented. And to sort of discover properties of them and figure out how to build around them and how to deliver them to one of the largest areas of knowledge work in the U.S.

18:44Matt Harpe:is just methodologically very exciting. And one of the things that I think makes Basis an exciting place to be a part of is we really want to be the home for applied AI talent in New York. And I think there are a number of properties of accounting as a discipline that has forced us to be on the frontier of what's possible in terms of leveraging the abilities of these models to build very capable systems. And so I think those two things sort of go in tandem. And to me, there are a few things that are more exciting to be working on than the confluence of those two things. Well, I'm glad you have such high conviction because there are some cool AI startups also here in New York that you're competing for talent with.

19:23So I guess you have to sell the vision there versus Ramp AI or Modal or some other really interesting companies here. And I guess I'm curious, when you were hiring early on, was an interest in accounting important for these people? Was it about, hey, we're going to grow really fast. This is a great place to build a career. What was the sort of elevator pitch to get that top person?

19:46Matt Harpe:One of the most important things I think we've done as a company is make sure that Basis is a home for exceptional ML researchers, engineers, etc. And pretty much none of them have any background in accounting. It's very rare to find someone like that who has a background in accounting. There aren't just people at OpenAI looking to join a new startup who also are accountants. You might think. You might think. Very few of them actually have engineering or ML experience, I would say, by background. There is like a probably 5-10 % of the company who have no background in accounting, but whose parents are accountants.

20:20Matt Harpe:So I think that is actually an interesting observation, but that is a relatively small portion of people. I think the thing that we talk to folks about, and we still talk to folks about today, is because we didn't start by building a chatbot. I think pretty much all, most of the sort of exciting applied AI companies to date have been focused on this sort of text-in, text-out chat experience. But because we started building around these agentic systems from day one, because we thought that was what was required to be successful in moving the needle in accounting, we were forced to, I think, do a lot of interesting work around how to build these agentic systems early on, build, I think, a pretty strong competency around that.

20:57Matt Harpe:And we've been able to compound around that for quite a while, to the point where I think we have among the most interesting and capable agents in production of really any applied AI company, in part because of the preconditions that existed in accounting, and then in part because of how we've chosen to focus. And so I think for a lot of people who are really interested in pushing the bounds of what's possible within sort of the broader applied ML space, Basis is a pretty exciting place to work. And I think of places that are headquartered in New York and have all of their core research and engineering efforts in New York City.

21:33Matt Harpe:I think potentially an end of one company in that respect. And so that's, I think, a big part of the reason that folks joined us early on and continue to join us today. These days, you can chat with AI about almost any business problem. Rippling AI is built to actually solve them. That's because Rippling AI is built on your live workforce data. that gives you full visibility into your startup and the ability to take action across every department. Say you're planning your next few hires. Just ask Rippling AI, what would it cost to add two more engineers this quarter? You'll instantly see a breakdown of comp ranges by level and location.