Rillet’s $100M Raise Is a $1B Warning to Oracle and Intuit
The AI winners won’t be the companies with the flashiest chatbot. They’ll be the ones trusted to touch the money — and Rillet just put $100 million behind that bet.
Most businesses are using AI to write emails while their finance team is still babysitting software designed before the iPhone. That is not innovation. That is putting a turbo on a fax machine.
Rillet, an AI-native accounting platform, has raised a $100 million Series C at a $1 billion valuation. More interestingly, founder Nicolas Kopp says the round came together in 48 hours after the company showed existing investors its momentum.
That should make the executives at Oracle, Intuit, SAP, Workday and NetSuite a bit uncomfortable. Not because Rillet has magically won accounting. It hasn’t. But because it is attacking the part of business software where AI can produce actual dollars: the machinery that closes the books, recognises revenue, reconciles transactions and tells an owner whether the business is making money or merely creating a very expensive hobby.
A $100 million vote against legacy ERP
Rillet announced the raise on August 17, 2026. ICONIQ led the round, with Sequoia, Andreessen Horowitz, Bain Capital Ventures and others participating. It brings the company’s total funding above $200 million.
The headline number is flashy. The operational numbers matter more.
Rillet says it has more than 600 customers, doubled new annual recurring revenue in the three months before the raise, and is expanding beyond its original tech-and-AI customer base into biotech, healthcare, fintech, logistics and professional services. Fortune reported that roughly 40% of its customers now sit outside tech and AI.
That is the tell.
Every software company says it can serve everyone. Most are lying, or at least getting ahead of themselves. But when a finance product starts moving from fast-growing AI companies into recycling businesses, movie studios and ordinary operating companies, it is no longer just selling to people who enjoy playing with new toys. It is being tested where mistakes have consequences.
Rillet’s pitch is not that accounting should become a chatbot. Thank God. A chatbot that confidently invents a number is amusing when it writes a bad LinkedIn post. It is catastrophic when it files a tax return or closes a quarter.
The pitch is that the general ledger, the system of record beneath the finance function, should be built for AI agents from the start. Those agents can propose work across journal entries, reconciliations, revenue recognition, reporting and close management, while people approve the consequential decisions and retain an audit trail.
That is a much more serious proposition than sprinkling a generative-AI button on top of 20-year-old enterprise software.
Why accounting is a far better AI target than content
The market has become obsessed with AI that talks. That is understandable: it is visible, entertaining and easy to demo. But conversational AI is rapidly becoming a commodity. Plenty of models can produce decent text. The value is moving into the systems, workflows and proprietary data around them.
Accounting is almost perfectly built for that shift.
It is repetitive, rules-heavy, data-rich and painfully expensive when done badly. It also has a clear definition of success. Either the ledger reconciles, the revenue treatment is defensible and the audit evidence exists — or it doesn’t. There is far less room for the usual AI nonsense where a vendor claims “productivity transformation” because someone saved six minutes drafting a meeting summary.
Rillet says its platform lets customers choose among foundation models and prevents those models from training on customer data. It also says it has built governance tools that show accountants the actions an AI agent took, the numbers it used and how it reached an outcome.
That governance layer is not a footnote. It is the product.
A business owner should never ask, “Can the AI do this?” The grown-up question is, “Can I see exactly what it did, stop it when it is wrong, and defend the result to an auditor, regulator, lender or buyer?”
If the answer is no, you do not have an AI system. You have an intern with no supervision and infinite confidence.
The overlooked opportunity is the accountant shortage
There is a lazy story doing the rounds that AI will wipe out accounting jobs. The evidence does not support the simple version of that story.
The US Bureau of Labor Statistics projects employment of accountants and auditors will grow 5% between 2024 and 2034, adding 72,800 jobs. It expects about 124,200 openings a year over that decade. The agency’s view is that automation will reduce routine work while making advisory and analytical work more important.
That sounds right to me.
Good operators know the bottleneck is rarely “we have too many talented people doing dull manual work.” The bottleneck is that your best people are buried in dull manual work because you do not have enough of them.
Kopp has pointed to an accountant shortage as part of Rillet’s demand story. That is more credible than the usual AI promise of mass replacement. A finance team with two excellent people does not need a robot CFO. It needs fewer hours wasted chasing invoices, fixing data-entry errors, copying figures between systems and rebuilding evidence for the audit.
Give competent people leverage and they can do the work that actually improves a company: pricing decisions, cash forecasting, margin analysis, acquisitions, capital allocation and spotting the dodgy assumption before it becomes a nasty surprise.
That is why AI-native finance software could matter far more than AI-generated marketing copy. It does not just make the business sound busier. It can make the business run better.
Oracle and Intuit have a distribution advantage — until they don’t
Let’s not get carried away. Oracle, SAP, Intuit and their peers are not defenceless relics waiting politely to be disrupted. They have entrenched customers, deep integrations, armies of implementation partners and enormous budgets. Replacing an ERP system is a major decision. It is difficult, risky and often deeply unpleasant.
That is precisely why Rillet’s traction is worth watching.
According to Kopp, about half of Rillet customers come from Intuit, 30% from NetSuite and Sage Intacct, and the remaining 20% from systems including Oracle, SAP, Workday and Microsoft. Those are company claims, not independently audited market-share figures, but the direction is what matters: customers appear willing to rip out core financial software when a replacement offers a material improvement.
Rillet has also formed an alliance with EY. Again, this is not proof of victory. But it addresses the real objection: finance software cannot simply be clever; it must be controlled, auditable and implementable in the messy real world.
The old ERP vendors face an awkward problem. They can add AI features, and they will. But retrofitting autonomy, real-time data flows and auditability into a legacy architecture is not the same as building for them from day one.
A heritage building can be beautifully renovated. It is still harder to turn it into an airport than to build an airport.
The contrarian take: don’t buy “AI-native” software yet
Here is where founders and operators can make an expensive mistake: seeing Rillet’s raise and deciding to replace every core system with an AI-native alternative next Tuesday.
Don’t.
A $1 billion valuation is not a warranty. A $100 million funding round is not a clean audit. And AI agents should not be allowed to touch payments, payroll, tax filings, revenue recognition or customer credits without clear approval limits, tested controls and an accountable human owner.
The lesson is not “buy the newest platform.” The lesson is that your current finance stack deserves a brutally honest review.
Ask whether your systems create a reliable real-time view of cash, gross margin, revenue, liabilities and working capital. Ask how many people touch a number before it reaches the board pack. Ask whether your month-end close is a controlled process or a heroic ritual performed by exhausted people with too many spreadsheets.
If the answer makes you squirm, that is useful. Businesses improve when the truth becomes too uncomfortable to ignore.
What this means for you
If you are a founder, operator or investor, do three things this week.
First, measure the cost of finance friction. Do not ask your CFO for a vague view on “efficiency.” Ask for the number of days to close, the number of manual reconciliations, the number of spreadsheet handoffs and the cost of errors or rework. You cannot improve a fog.
Second, pick one high-volume, low-discretion workflow for an AI pilot. Think invoice coding, bank reconciliation preparation, expense-policy checks or variance explanations. Keep a human approver. Define a baseline: hours saved, error rate, turnaround time and exceptions escalated. If the vendor cannot show you a full audit trail, walk away.
Third, make your finance team the owner, not IT’s guinea pig. The people who carry the liability for the numbers must design the controls. Your technology team should help with integration and security; it should not be deciding whether an AI entry is acceptable in the general ledger.
Rillet’s $100 million raise is not really a story about another unicorn. It is a warning that AI is moving from the front office, where it helps people look productive, into the back office, where it can alter margins, speed and decision quality.
That is where the serious money will be made. And where the careless operators will get found out.