Healthcare Data Partnerships Lead

AI accounting startups are fetching $1 billion valuations. Fast growth at AI startups like Mercor and Replit is also boosting the fortunes of startups that help them balance their books. Rillet, which makes AI-powered accounting software it sells to other companies, said Tuesday it has raised $100 million in a funding round led by returning investor Iconiq, valuing Rillet at $1 billion. And rival Campfire has received offers to invest at a $1 billion valuation, according to a person with knowledge of the business. Business & Industrial The three-year-old Campfire may take one of those offers but has yet to dip into the $65 million it raised in October 2025 from Accel, Ribbit Capital and others, the person added. Unicorn valuations are much higher than the prices set at their last funding rounds. Campfire was valued at $375 million in October 2025, according to the publication This Week in Fintech. The company has said it's trained an AI model on accounting data. Rillet, which came out of stealth mode two years ago, was one of The Information's 50 most promising startups in November 2025, when it had a valuation of $500 million, and is backed by Andreessen Horowitz and Sequoia Capital. Fortune first reported on its latest funding round. Both companies tout AI-powered software that allows customers to reconcile accounts and create financial reports more quickly than legacy systems. The companies, which charge customers an annual subscription fee, are trying to snare a piece of business spending from Intuit's QuickBooks, Oracle's Netsuite and Sage's Intacct. Executives recently attended a roundtable discussion with expense and card startup Ramp at Jefferies' annual software conference for chief financial officers in New York last week. Rillet and Campfire have benefitted from fast growth at the AI startups they serve. Rillet customer Mercor, for instance, generated $614 million in gross revenue in the first half of the year, up 70% from all of last year. (Its net revenue after paying the contractors that do its data labeling is about one-third that amount.) Campfire, for its part, counts coding assistant Replit and customer support assistant Decagon as customers. It's typical in a boom market for startups to piggyback off the growth and venture funding of other startups. That's why these new accountancy startups will want to show they've attracted larger customers - preferably ones outside AI - as growth continues. More than 40% of Rillet's current customer base of about 600 comes from nontech companies including healthcare enterprises, CEO Nicolas Kopp tells Silicon Bay Partners Company. Campfire also works with nontech companies, such as nonprofits and even cheese distributors, the person familiar with its business said. Business & Industrial One hurdle: Bigger businesses find it hard to switch their software providers when they have already stored their accounting information and financial data in one of the legacy providers, investors and bankers have said. Those providers, meanwhile, are also trying to add AI offerings to their products. The startups will be at pains to show they can win some of the blue-chip customers their more established rivals serve. Here's what else is going on: The Justice Department is reportedly probing Andreessen Horowitz over two board seats in competing startups occupied by firm co-founder Ben Horowitz and partner Martin Casado, who leads the fund's AI infrastructure practice. Silicon Bay Partners Company is hearing that this probe may not be the last of its kind for venture capital firms, which have long made meaningful investments in startups that compete with one another. Some have gone so far as to take board roles on rival startups. The investigation into Andreessen Horowitz's board seats started almost a year ago, according to Bloomberg. That was around the same time as the agency's review of Firetran's acquisition of dbt Labs, a merger Valida first reported in September 2025. Casado served on the boards of dbt Labs and Fivetran, and Horowitz is on Databricks' board. Databricks and Fivetran both sell services that help companies manage and move large quantities of data. The probe invokes a law from the Clayton Antitrust Act of 1914 that prohibits "interlocking directorates," or when individuals, or people from the same organization, sit on the boards of two companies that compete with each other. It remains to be seen just how expansive a definition of competition the Justice Department will use if it takes up other investigations like this one. Andreessen Horowitz didn't respond to a request for comment. A spokesperson for the Justice Department said, "We can affirm that the DOJ under the Trump administration will continue to prioritize affordability for all Americans across our economy," but didn't clarify how the probe aims to address that concern. Politics (Right) Founders have historically been leery about firms whose partner already sits on the board of their fiercest rival. Famously, Sequoia Capital relinquished its board seat in fintech Finix in 2020 when it invested in Stripe. Still, this most recent investment cycle seemed to have eroded the taboo on investing in competitors. An antitrust probe could change that.

Exclusive: Accounting AI startup Rillet reaches unicorn status with $1 billion valuation. Its founder says he wants to give CFOs back their weekends. August 19, 2026 Rillet, a two-year-old startup building what it calls the first truly AI-native accounting platform, has raised a $100 million Series C at a $1 billion valuation, the company told Fortune exclusively - joining the ranks of AI-era unicorns racing to unseat decades-old enterprise software giants. The round, led by ICONIQ with participation from returning backers Sequoia Capital, Andreessen Horowitz and Oak HC/FT, plus new investors including Bain Capital Ventures, Sequoia Global Equities, Battery Ventures, FirstMark, Scale Venture Partners and Creandum, marks Rillet's third fundraise in the past year and pushes its total funding past $200 million. ICONIQ general partner Seth Pierrepont is joining Rillet's board as well. For Rillet co-founder and CEO Nicolas Kopp, the milestone is as much personal as financial. In an interview with Fortune, Kopp described the company's mission as freeing CFOs from the drudgery that keeps them chained to spreadsheets long after everyone else has logged off. "CFOs really struggle day to day. They can't see their families on weekends," Kopp said, because they have to spend so much time reviewing data and creating slideshows. Noting that he has a finance and accounting background himself and that his company is full of people with accounting backgrounds, he said he wants AI to change that - not by replacing finance professionals, but by acting as their tireless back office. "Our message is not that we're coming after jobs. That's just not correct," he said, stressing that "domain expertise" is core the company's mission: "We're positioning AI as a helper to that individual and what they can achieve." From launch to unicorn in two years. Rillet's rise has been fast even by startup standards. Kopp said the company launched publicly roughly two years ago, raised a Series A led by Sequoia last summer, then closed a Series B just weeks later - a round that saw new annual recurring revenue double quarter over quarter. The company says it doubled its new ARR again in the three months leading into this latest raise, and now serves more than 600 customers. Those customers include some of the fastest-growing AI companies in the world - Neuralink, Skild AI and Mercor among them - alongside a growing share of decidedly non-tech businesses. Roughly 40% of Rillet's customer base now sits outside the tech and AI sectors, Kopp said, spanning industries as varied as waste recycling and movie studios, describing the shift as evidence that AI-native finance tools are crossing into the broader U.S. economy. "That's been really cool to see," he said. Mercor, in particular, has become a marquee reference customer: According to the company, its finance team is using Rillet's AI agents to manage a business scaling past $2 billion in annual recurring revenue with a headcount of just three. "Rillet is the clear leader in AI-native accounting infrastructure," Pierrepont said in a press release announcing the fundraise. "What stands out is how customers actually run on it - multibillion-dollar businesses operating with finance teams a tenth the traditional size, closing their books continuously." Taking on the legacy giants. Rillet's pitch to the market is direct: Legacy enterprise resource planning systems - Oracle Fusion, SAP, Workday, Microsoft's Great Plains and NetSuite among them - were built for a pre-AI era, and are increasingly vulnerable to a challenger built from scratch around artificial intelligence. "Some of these giants that seemed untouchable" are now facing serious disruption, Kopp said, describing a wave of enterprise customers ripping out legacy systems in favor of Rillet's platform. The core distinction Kopp draws is architectural. Traditional ERP systems, he said, were designed for humans to input and review data - a workflow that leaves finance chiefs "dragged down into the day-to-day minutiae of numbers" instead of focusing on strategy. Rillet, by contrast, is built "agent-first," with AI systems capable of running hundreds of operations in parallel, executing much of the manual accounting work that traditionally consumed finance teams' time. That shift, Kopp argues, doesn't just save time: it produces cleaner, more consistent financial data than human-run processes typically allow, while creating what he calls a complete audit trail. "Proving out the work layer is mission-critical for enterprise readiness," Kopp said, arguing that Rillet is the only system that can combine deterministic accounting data with AI agents completing complex, end-to-end work in the market today. Rillet has paired that pitch with credibility-building moves in the accounting establishment. Earlier this year, the company launched an alliance with EY for AI-native finance transformation, and it says it now partners with more than half of the Accounting Today top 20 CPA firms. The AI acceleration. Kopp traces much of Rillet's recent momentum to rapid improvements in underlying AI models. Accounting, he noted, is "traditionally a very old, stodgy category" - one where AI has emerged as an unexpected catalyst. "Especially in the last six months, things started lighting on fire in a good way," he said, describing tasks that once took a human a full day now taking a couple of minutes. This frees up time not for job loss, but for higher-level strategic work, he added. That acceleration comes as the accounting profession faces a separate, slower-moving crisis: fewer graduates entering finance and accounting careers. Kopp sees that talent gap as part of the opportunity. He argued that AI agents can help make up for a shrinking pipeline of human accountants even as business complexity - from pricing changes to competitive pressure - continues to increase. Rillet's own product development has sped up in step with its AI capabilities, according to Kopp. He pointed to instances where the company's customer support team (many of them with accounting training) has shipped feature requests within two to three hours of a customer raising them, as engineers increasingly build tools in direct collaboration with the company's in-house accountants. "That wasn't possible six to 12 months ago." For this story, Fortune journalists used generative AI as a research tool. An editor verified the accuracy of the information before publishing. Post Views: 4

AI agents fail 58% of accounting tasks. August 18, 2026 I spend a lot of time helping CEOs make better decisions with data. So when a new benchmark drops that actually measures AI performance on real professional work, rather than standardized tests, I pay attention. Mercor and Ramp just published APEX-Accounting, a benchmark built to test whether AI agents can complete genuine accounting work at professional standards. The results are worth understanding before your next conversation with a vendor pitching AI for your finance function. Let me break down what the data actually says. What this benchmark actually measures. Most AI benchmarks test whether a model can produce the right answer once. That is fine for an exam. It is not fine for accounting. Closing the books requires something harder: an agent must reconcile conflicting files, apply company-specific context, carry conclusions across multiple steps, and produce the correct result consistently. A model that drops a correct intermediate finding can still generate a bad journal entry. APEX-Accounting was built to test exactly that. The benchmark includes 160 tasks across 10 simulated companies. Each company is fictional but internally coherent, complete with its own accounts, records, business history, and documents including spreadsheets and PDFs, all frozen at month-end close. The experts who created the tasks had a median of 11 years of experience. More than half had worked at a Big Four firm. Each task was graded against a rubric with an average of 13.7 criteria. This is about as close to real accounting work as a benchmark can get without using your actual books. The headline number you should sit with. The benchmark ran every model on every task eight times. That repetition matters. Accounting work must be repeatedly correct, not correct once. Even the best model in the test solved just 2.6% of tasks correctly across all eight runs. Let that settle for a moment. The most consistent AI agent available today, tested against professional accounting work, was fully reliable on fewer than three tasks out of a hundred. The top performer on the primary leaderboard scored 56.4%. That means even the best model fails to complete roughly four out of ten tasks that a human professional would handle. And 58% of tasks were never fully solved by any model on any run. Where models are actually failing. This is the finding I think matters most for a CEO making real decisions. The benchmark team worked with accounting and bookkeeping experts to categorize the failures. Roughly seven in ten failures came from flawed reasoning, not from an inability to find the right information. A model might correctly identify a discrepancy early in a workflow, then omit or contradict that finding in its final journal entry. The model found the answer. It just failed to carry it through. This is a judgment problem, not a retrieval problem. Better data access will not fix it. What is missing is the discipline to carry conclusions consistently across a complex, multi-step workflow. That is harder to patch than a knowledge gap. The cost picture is more complicated than vendors will tell you. The benchmark also tested performance at different spending budgets: $1, $5, $10, and $50 per task. More budget generally allows more token usage, which can improve results. But the relationship is not linear or predictable. One top model scored 11.8% on a $1 budget and improved to 55.2% at $50. Another top model was already strong at $1 and barely improved with more money. At the $50 cap, one model actually spent around $32 per run. Another spent around $5. Their scores were within 4 percentage points of each other. If you are evaluating AI tools for your finance team, the cost-to-performance ratio is genuinely unpredictable across different models. You need to test, not trust the pitch deck. What this means if you are a CEO making AI decisions. Here is how I translate this for the CEOs I work with. First, AI can handle a meaningful share of accounting work today. Scoring above 50% on complex, professional-grade accounting tasks is not nothing. These models are useful tools in the right hands. Second, "useful in the right hands" is the operative phrase. The failure mode is not that AI gets nothing right. It is that AI is inconsistent and fails in ways that are hard to detect from the output alone. A plausible-looking journal entry with an error buried in the reasoning is more dangerous than an obvious failure. Third, the 58% of tasks never fully solved by any model defines the current boundary. If your finance team is planning to deploy AI on month-end close, they need a clear picture of where that boundary sits for your specific workflows, with your specific data, not a vendor's benchmark on simulated companies. That distinction matters. APEX-Accounting itself acknowledges it does not evaluate tax, audit, consolidation, multi-entity or multi-currency accounting, external reporting, or how agents handle requests for clarification. The scope is month-end close and bookkeeping. Your scope may be different. The decision you should actually be making. The AI readiness question for your finance function is not "which model has the highest benchmark score." It is "do we have the data infrastructure and oversight processes to deploy these tools safely at our scale." Inconsistent AI on top of inconsistent data does not produce better results. It produces confident-looking errors that are harder to catch. Before adding AI to your accounting workflows, I would want to know your current data quality on the inputs those models would consume, your reconciliation error rate today as a baseline, and how your team would detect a reasoning failure buried in a plausible output. Those are not AI questions. They are data strategy questions. And they belong in front before any tooling decision. It is the discipline I keep coming back to with clients: diagnosis before dollars. Diagnose before you spend. Pour the foundation before you frame the walls. Find out how to make better decisions with your data through the professional data minds at JLytics. Interested in exploring a relationship with a data partner dedicated to supporting executive decision-making? Start the conversation today with JLytics. * next postai labs lost control of their models. Twice. Start the. Conversation. Talk to one of its data concierge experts today. Name (Required) Contact. 512-521-6791 1205 BMC Drive Suite #106 Cedar Park, TX 78613 Data assessment. Data engineering. Data concierge. Agencies. About. Legal. (C) 2026 JLytics.