Basis is an AI agent platform for accountants.
Processing details
site,
Matt Harp, co-founder and CEO of Basis, said: today’s accounting About how his multibillion-dollar startup pays homage to the accounting profession by naming its New York office conference rooms after accounting and finance pioneers like Luca Pacioli and Alexander Hamilton.
“There’s a lot of talk out there about AI and people are very excited, and I think it’s no surprise,” Harp said. His company provides AI technology for accounting, tax, and auditing, and he estimates that its technology is deployed in about 40% of the top 25 companies and more than 25% of the top 150 companies in the United States. Earlier this month, Chicago-based top 50 company Sikich
“As far as we know, we are the most deployed AI platform in accounting firms,” says Harpe. “Our approach is to try to bring together both great accountants and some of the strongest machine learning researchers and engineers in New York.”
Businesses use it in accounting, tax, and auditing practices. “All of these companies are actively using it in at least one area of their business,” he added. “We don’t participate in every practice area of all these companies. In general, the way we approach it is that we want to partner with them to help companies achieve the most AI-native practices possible. We’re really only working with companies that are implementing it across their practices. It’s not something that one person or 10 people can use. In all of these practice areas, we basically see Basis as the core of their AI strategy.”
New York AI
The company, with all 120 employees working out of its New York office, competes for talent with AI giants like OpenAI and Anthropic.
“Our goal is to become the home of applied AI in New York,” Harpe said. “We have amazing engineers and researchers in New York, and we’re very lucky that many of them chose Basis. They want to take on the challenges at the forefront of applied AI. They don’t just build another chatbot or bolt something onto an existing product, they actually figure out how to build the most sophisticated agent systems that are accurate enough to be used in real-world production workflows. “The other thing is, they want to contribute to doing real work in the real economy. I think people are excited that these agents are actually doing real things. “If you look at the number of companies that are doing this, most of them are not headquartered in New York, but we’re working with AI in all areas, not just accounting. I think we are one of the few organizations that are on the cutting edge of building systems. ”
All of Basis’ R&D activities are done in New York, not Silicon Valley like many AI companies. “Obviously, we’re not as big as OpenAI or Anthropic from a headcount standpoint, so there’s an opportunity for people to have real responsibility in building these systems instead of being one of many,” Harpe said. “It’s attractive from that perspective.”
Unlike those AI giants, his company focuses on the accounting profession. “Our mission is to figure out how we can actually make AI work as accurately as possible in accounting, which ultimately will be useful,” Harpe said.
Making AI trustworthy enough for accounting
He believes that AI can definitely automate the tasks of accountants. “Our real goal is to give every accountant in a company this AI that goes out and takes over the most manual parts of the job and basically reports back to the accountant so they can get those tasks completed on an ongoing basis,” Harpe said. “I think we have a strong track record of actually delivering results. We can trust the AI because it is accurate enough and capable enough to handle these complex workflows with a high level of precision.”
Generative AI’s large-scale language models have been accused of being “illusory” in creating seemingly plausible information from vast amounts of data siphoned from the internet and social media. Harp believes the model is now being improved.
“I actually think the trust part of accounting is very important,” he added. “Accounting is all about trust. That’s a core part of what this profession is dedicated to. You want to codify and structure all the economic activity that happens in and around these organizations and know that that integration can be trusted. That’s what accounting does, so trust is very important. And I think most AI companies don’t really focus enough on accuracy. What we prioritize first and foremost is the power and accuracy of the AI.”
Basis has a Deployed Intelligence team that helps accountants further trust in their AI by demonstrating its accuracy.
“We use this program to run workflows that they know well and know inside and out,” Harpe says. “We get them to start working on these workflows with AI so they can see first-hand what AI can and cannot do. This builds intuition and trust. It’s like bringing someone new to your team on board: give them a few things, see how they’re doing, and then move on to the next thing. We help them build trust as part of that process.”
A big part of that trust involves preventing sensitive client information from being sucked into publicly available AI systems.
“At the end of the day, it also comes down to trust,” Harp said. “Clients trust businesses. That relationship is very sacred and we don’t want to end up putting that data in a place where it’s going to be exposed in any way. Having a very clear security policy and a clear AI policy is really important because you can basically guarantee to the business that this data is completely siled. In fact, we don’t just silo data by company; we actually silo data by client, so… You can’t sile data that resides across one client.” Companies are becoming more sophisticated about what policies are in place when working with partners to ensure data is used appropriately. ”
He discovered that companies have detailed surveys that they ask technology providers like Basis. How we handle certain situations and what cybersecurity certifications we hold.
“People are definitely becoming more sophisticated and making sure that proper controls are in place,” Harp said. “We’re very focused on ensuring data integrity is maintained. You can’t train data on public models, and you can’t potentially transfer data from one client instance to another.”
The impact of AI on employment
Many accountants are concerned about losing their jobs to AI, and while this tool is useful from an efficiency standpoint, it can also be perceived as a threat.
“I strongly believe there is still a lot of accounting work to be done in the world,” Harp said.
He cited his past experience developing machine learning systems for the medical industry. “My understanding is that there are no hospitals in America that can actually tell you how much it will cost to perform a procedure on a patient,” Harp said. “If you’re going to get an MRI scan in a hospital, the hospital doesn’t know how much it’s actually going to cost. That’s a problem, because if you’re a hospital, one of your main objectives should be how to reduce the cost of long-term care. Either increase your profits or provide more cost-effective care to your patients. The fact that you can’t do that is an accounting challenge in my opinion. Accountants’ job is to account for all the economic activity that happens in and around the hospital. There’s a lot of other things you can’t get done if you spend all your time focusing on compliance activities, but I believe that if you can spend time on these very manual aspects of the job that everyone is thinking, “Why can’t I do this on a computer?” people will have more time to do things, and we could actually end up with more accounting jobs within five years. ”
He compared this to the impact of AI on software engineering and explained that such employees are still very much needed. “Software engineering is ahead of accounting in terms of AI adoption,” Harpe says. “At our company, we don’t have any engineers who write any meaningful amount of code anymore. Very few of us write code. But, of course, it used to be a core part of what they did all day. But our engineering and machine learning teams are as busy as ever. The reality is that we’ve gotten better at people, so we can now build things that we wouldn’t have thought to build before.”
When asked what the company is building with the funding it has received so far, he said most of it will go toward hiring and computing costs. Initially focused primarily on outsourced accounting technology, Basis has since expanded into tax and audit, and is earmarking the funds to continue building a team that can support businesses across the country in these areas. Basis has received more than $130 million in funding to date from investors including Accel, Khosla Ventures, Google Ventures, BTV, NFDG, Box Group, and former Goldman Sachs CEO Lloyd Blankfein. They’re looking beyond the recent volatility in tech stocks.
“We have investors who are really long-term oriented, so we’re focused every day on building the long-term that we think is important, not the day-to-day,” Harp said.
Basis’ Ode to Accounting website is also long-term and relies on:
“One of the things we’re very focused on is making it clear that accounting is actually very important and that accounting will play a very important role in the coming AI era,” Harpe said. “We’ve been working on ‘Ode to Accounting’ for the better part of a year, and it’s all about the history of accounting and how we got here. Accounting was the first written product by humans. In our opinion, all Underpinning how a market economy works is the ability to understand what’s going on in the detailed decisions of each organization. Sometimes experts lose sight of that, but in the next era, as economies become more complex, it will become a reality.” It is important to recognize that accounting is central to providing the readability to engage in more complex forms of economic interaction, such as those in which AI interacts with each other. ”
