Marble, a startup constructing synthetic intelligence brokers for tax professionals, has raised $9 million in seed funding because the accounting {industry} grapples with a deepening labor scarcity and mounting regulatory complexity.
The spherical, led by Susa Ventures with participation from MXV Capital and Konrad Capital, positions Marble to compete in a market the place AI adoption has lagged considerably behind different information industries like legislation and software program growth.
"Once we appeared on the economic system and requested ourselves the place AI goes to rework the way in which companies function, we targeted on information industries — particularly companies with hourly fee-based service fashions," stated Bhavin Shah, Marble's chief government officer, in an unique interview with VentureBeat. "Accounting generates $250 billion in fee-based billing within the US yearly. There's an incredible alternative to extend effectivity and enhance margins for accounting companies."
The corporate has launched a free AI-powered tax analysis device on its web site that converts advanced authorities tax knowledge into accessible, citation-backed solutions for practitioners. Marble plans to develop into AI brokers that may analyze compliance eventualities and ultimately automate parts of tax preparation workflows.
Marble's backers share Shah's conviction in regards to the market. "Marble is rethinking the accounting system from the bottom up. Accounting is likely one of the greatest — and most neglected — markets in skilled providers," Chad Byers, basic companion at Susa Ventures, instructed VentureBeat. "We've recognized Bhavin from his time as an government within the Susa portfolio, and have seen firsthand how sharp and execution-driven he’s. He and Geordie convey the right mixture of operational depth and product intuition to an area lengthy overdue for change — they usually see the identical large alternative we do."
The accounting {industry} misplaced 340,000 staff in 4 years — and replacements aren't coming
Marble enters a market formed by structural forces which have essentially altered the economics {of professional} accounting.
The accounting occupation has shed roughly 340,000 staff since 2019, a 17% decline that has left companies scrambling to fulfill shopper calls for. First-time candidates for the Licensed Public Accountant examination dropped 33% between 2016 and 2021, in line with AICPA knowledge, and 2022 noticed the bottom variety of examination takers in 17 years.
The exodus comes as child boomers exit en masse. The American Institute of CPAs estimates that roughly 75% of all licensed CPAs reached retirement age by 2019, making a demographic cliff that the occupation has struggled to handle.
“Fewer CPAs are getting licensed 12 months over 12 months," Shah stated. "The {industry} is compressing on the similar time that there's extra work to be performed and the tax code is getting extra sophisticated."
The Nationwide Pipeline Advisory Group, a multi-stakeholder physique fashioned by the AICPA in July 2023, launched a report figuring out the 150-hour schooling requirement for CPA licensure as a big barrier to entry. A separate survey by the Middle for Audit High quality discovered that 57% of enterprise majors who selected to not pursue accounting cited the extra credit score hours as a deterrent.
Current legislative modifications mirror the urgency. Ohio now gives options to the 150-hour requirement, signaling that states are keen to experiment with pathways that might reverse enrollment declines.
Why AI remodeled legislation and software program growth however left accounting behind
Regardless of the occupation's challenges, AI adoption in accounting has moved extra slowly than in adjoining information industries. Harvey and Legora have raised a whole lot of thousands and thousands to convey AI to authorized work. Cursor and different coding assistants have remodeled software program growth. Accounting, against this, stays largely depending on legacy analysis platforms and handbook processes.
Geordie Konrad, Marble's government chairman and a co-founder of restaurant software program firm TouchBistro, attributes the hole to how folks conceptualize AI's capabilities.
“It was apparent to many individuals that LLMs may do significant work by manipulating code for software program builders and manipulating phrases for attorneys. Within the accounting {industry}, LLMs are going for use as reasoning brokers," Konrad stated. " That requires a bit extra of a two-step evaluation to see why it's a giant alternative."
The technical problem is substantial. Tax laws kind one of the crucial advanced, interconnected info techniques that people have created — tens of hundreds of interlocking guidelines, steering paperwork, and jurisdiction-specific necessities that continuously overlap or battle.
"If you wish to put AI by means of its paces and ask how far it's are available replicating cognitive features, that is an unbelievable playground to work in," Konrad stated.
A dramatic shift: AI adoption amongst tax and finance groups doubles in a single 12 months
Current knowledge suggests the accounting occupation's stance towards AI is shifting quickly.
A 2025 survey from Hanover Analysis and Avalara discovered that 84% of finance and tax groups now use AI closely of their operations, up from 47% in 2024. The 2025 Generative AI in Skilled Companies Report from Thomson Reuters Institute discovered that 21% of tax companies already use generative AI expertise, with 53% both planning to undertake it or actively contemplating it.
Massive accounting companies have invested closely in AI infrastructure. Deloitte has developed generative AI capabilities inside its audit platform. BDO introduced a $1B funding in AI over the following 5 years. EY launched an AI platform combining expertise with technique, transactions, and tax providers. PwC estimates an entire AI-driven audit answer will launch by 2026.
However adoption at smaller companies stays uneven. In response to Thomson Reuters analysis, 52% of tax agency respondents who use generative AI depend on open-source instruments like ChatGPT moderately than industry-specific options—a sample that might shift as purpose-built options emerge.
Marble's founders consider the hesitance stems not from technophobia however from an absence of compelling choices.
“Corporations need to embrace AI," Shah stated. “They simply haven't seen nice software program and tooling made for them. That's a part of the chance — to work with them and construct one thing they're excited to make use of on a day-to-day foundation.”
Can synthetic intelligence rescue accounting's billable-hour enterprise mannequin?
AI's arrival in accounting raises questions in regards to the occupation's billing construction.
Accounting companies have historically generated earnings by billing shoppers for employees time, usually at multiples of worker compensation prices. Junior associates performing compliance work signify a big income stream. If AI can automate that work, does it undercut the enterprise mannequin companies rely on?
Marble's founders argue the alternative. The power staffing scarcity has already constrained companies' capacity to seize accessible income. Advisory and consulting work — higher-margin providers that shoppers actively need — goes undone as a result of practitioners are buried in compliance duties.
"Everybody within the {industry} agrees that an infinite quantity of advisory work merely isn't getting performed," Konrad stated. "Clients need it. Corporations need to do it as a result of it's high-margin, nice work. However no person will get to it."
The 2025 AICPA Nationwide Administration of an Accounting Follow Survey helps this view. Corporations reported a median 6.7% improve in internet shopper charges over the prior 12 months, with development in audit, assurance, tax providers, and shopper accounting advisory. Web remaining per companion climbed 11.9% from fiscal 12 months 2022 to fiscal 12 months 2024, reaching $252,663.
The survey additionally discovered rising curiosity in AI adoption, although most companies have but to allocate formal budgets or develop structured coaching packages. Continued adoption, the survey prompt, may assist develop providers and gas continued development.
Accountants gained't undertake AI instruments they’ll't belief with delicate shopper knowledge
For AI to achieve accounting, it should clear a excessive bar for knowledge safety. Accounting companies deal with a number of the most delicate monetary info within the economic system. Practitioners can’t undertake instruments that create compliance or confidentiality dangers.
In response to Avalara's survey, 63% of respondents cited knowledge safety and privateness considerations as the highest limitations to automating tax and finance features. The priority persists all through the adoption lifecycle, from preliminary choice by means of implementation and ongoing use.
Marble has made safety a foundational precedence. The corporate obtained software program compliance certification earlier than releasing any product and maintains that knowledge privateness is embedded in its operational tradition from day one.
"Safety is on the core of what we’re constructing," Shah stated. "Each worker is aware of that safety is crucial. It's part of our onboarding and one thing that we contemplate in every little thing we do."
From quantity crunchers to strategic advisors: How AI may reshape accounting careers
Marble's founders reject the narrative that AI will solely take away from accounting jobs. They suggest as an alternative that AI will lead to accounting jobs turning into extra strategic and fewer characterised by repetitive execution.
They draw an analogy to structure, the place computer-aided design changed laborious handbook drafting. Architects didn’t disappear — they gained instruments that permit them spend extra time on artistic design and fewer on mechanical copy.
"For those who take a number of the hours-intensive, much less artistic work out of what being a junior or intermediate accountant is, and also you substitute it with a job the place you're knowledgeable who’s being artistic, synthesizing concepts, and in a position to delegate plenty of duties to AI assistant platform options, you find yourself with an {industry} that's simply much more enjoyable to function in," Konrad stated.
The shift may additionally enhance shopper outcomes. When accountants spend much less time on compliance, they’ll make investments extra within the strategic advisory work that shoppers worth.
"Not solely does the work develop into extra pleasurable due to what you possibly can deal with, however that's additionally what your shoppers are going to worth extra from you," Shah stated.
The aggressive panorama: Marble faces well-funded rivals and legacy giants
Marble enters a market with formidable incumbents and well-funded rivals. BlueJ, a worldwide tax analysis platform, has raised over $100 million. Thomson Reuters, CCH, and Intuit have deep buyer relationships constructed over many years.
However the founders see alternative within the transition second.
"AI has modified what’s attainable within the {industry}," Shah stated. "We’re going to work with and combine with some expertise gamers within the {industry} and in addition compete with different gamers with new merchandise powered by AI. In some circumstances we’re going to neglect in regards to the present expertise answer for doing issues and return to the duty itself. We now have completely new technological capabilities — how would you design one thing from a clean canvas that works with people to perform that activity?""
The choice to supply a free analysis device displays Marble's go-to-market philosophy. By giving practitioners entry and not using a paywall, the corporate goals to construct belief and display functionality.
"It permits us to show a extremely compelling product that’s purpose-built to those who are anxious about use AI or query undertake it. Now they don’t have to consider buying one thing that’s cost-prohibitive once they don't know combine it into their workflow," Shah stated.
The $250 billion query: Can a startup rework how America does its taxes?
Marble's roadmap extends past analysis. The corporate plans to develop AI brokers able to analyzing advanced tax eventualities, figuring out compliance points, and ultimately automating vital parts of compliance workflows — all whereas retaining practitioners in management.
The founders body success not when it comes to disruption however rebalancing. Right now's tax work skews closely towards compliance, leaving the strategic advisory providers that shoppers crave — and that generate increased margins—perpetually undone. Marble's wager is that AI can flip that equation.
"Everybody desires it to look extra like compliance is completed easier, and also you spend time speaking about technique and planning," Konrad stated. "How do we modify that mix of compliance versus technique and planning to technique and planning first—with compliance as one thing that has been made dramatically easier?"
Whether or not Marble can execute on that imaginative and prescient stays to be seen. The corporate faces entrenched rivals, a occupation that has traditionally resisted technological change, and the inherent unpredictability of constructing AI techniques for high-stakes monetary work.
However the founders are betting that the {industry}'s demographic shift will speed up adoption in ways in which earlier expertise waves couldn’t. With fewer accountants coming into the occupation annually and shopper calls for solely rising, companies could have an elevated urge for food to embrace instruments that permit their remaining workers do extra.
"AI goes to alter each {industry} — in some circumstances in methods that can assist enterprise fashions and in some circumstances in methods that can problem them. We consider AI is finally going to make accounting companies’ companies higher and extra worthwhile and on the similar time finish shoppers will get higher providers at higher costs," Shah stated.
The accounting occupation, it appears, is about to search out out which facet of that equation it lands on.
[/gpt3]
