Product2 distinct publishers3 min readUpdated
The AI accounting startup raised $100mn led by ICONIQ and names Oracle, SAP, Workday, Microsoft and NetSuite as pre-AI relics. Its proof point is a finance team of three.
The Product Desk · Product desk

Compiled by The Product DeskSomething wrong?How this is made
Rillet has raised $100mn at a $1bn valuation in a Series C led by ICONIQ, disclosed to Fortune as an exclusive and reported by Nick Lichtenberg [1]. The interesting part is not the unicorn label but the architectural claim underneath it: that the general ledger itself gets rebuilt for agents, rather than having a copilot bolted onto an existing system of record [9][11].
Chief executive Nicolas Kopp's pitch names its targets directly. He argues that legacy systems from Oracle, SAP, Workday, Microsoft and NetSuite belong to a pre-AI era because they were designed for humans to key in and check data [8]. Rillet describes itself as "agent-first", with agents running hundreds of operations at once inside a real-time general ledger while accountants review the output instead of producing it [9]. "For the last two decades, the ERP has been treated as a system of record, a place to store what already happened," Kopp said. "In the AI era, it has to become the operating layer for what happens next." [10] Finance agents, he added, "need to work inside the general ledger", not merely pull data from it [11]. In practice, TechCrunch reports, that includes continuously pulling data from sources such as Salesforce and Brex [27].
The growth numbers are steep, and the two accounts of them are not identical. Rillet told Fortune it doubled its new annual recurring revenue in the three months before the round [12]; TechCrunch reports the company doubled ARR over the same period [13]. Those are different claims, and the difference matters to anyone modelling the business. The company says it now serves more than 600 customers [14], with agent activity growing about 70 percent month on month according to Kopp [15], and roughly 40 percent of customers outside tech, in industries from waste recycling to film studios [16].
The sales pitch leans hard on one account. According to the company, Mercor uses Rillet's agents to run a business scaling past $2bn in annual recurring revenue with a finance team of three [17], which works out to roughly $667mn of ARR per finance employee [29]. ICONIQ general partner Seth Pierrepont, who joins the board [4], described customers as "multibillion-dollar businesses operating with finance teams a tenth the traditional size, closing their books continuously" [18].
Kopp is running two messages at once. "Our message is not that we're coming after jobs. That's just not correct," he told Fortune [19], while also writing that "we believe in lean, impactful finance teams" [20]. He frames a shrinking profession as the opening, citing 340,000 fewer accountants in the US than five years ago [21].
Trust is the real gate on selling a replacement ledger, and Rillet has been buying it: an EY alliance launched earlier this year, plus partnerships with more than half of the Accounting Today top 20 CPA firms, KPMG and RSM [22]. It also raised fast. Kopp said on X the round came together in less than 48 hours and that the company was not planning to raise [24], following a $70mn Series B led by ICONIQ and Andreessen Horowitz and a $25mn Series A led by Sequoia [25]. Those three rounds total $195mn, short of the stated $200mn-plus, so earlier or undisclosed funding fills the gap [28].
Watch whether the ARR ambiguity resolves in favour of the stronger reading, whether non-tech customers keep gaining share, and whether Campfire, founded in 2023 with about $100mn from Accel and Ribbit and positioned as a modern NetSuite, forces price competition [23].
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
Rillet raised $100mn at a $1bn valuation in a Series C led by ICONIQ; the company disclosed the deal to Fortune as an exclusive and Nick Lichtenberg reported it.
Returning shareholders Sequoia, Andreessen Horowitz and Oak HC/FT joined the round; new investors include Bain Capital Ventures, Battery Ventures, FirstMark, Scale Venture Partners and Creandum.
The round takes Rillet's total funding past $200mn, and ICONIQ general partner Seth Pierrepont is joining the board.
Rillet sells what it calls an AI-native ERP; Nicolas Kopp and Stelios Modes founded it in 2021 and it launched publicly in August 2024.
Rillet calls itself "agent-first"; its AI agents run hundreds of operations at once inside a real-time general ledger, and accountants review the work rather than enter it by hand.
Kopp said: "For the last two decades, the ERP has been treated as a system of record, a place to store what already happened. In the AI era, it has to become the operating layer for what happens next."
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Financing firm, product outcomes unverified
The financing facts are corroborated by two independent publishers and internally consistent apart from a ~$5mn gap between itemised rounds and the stated total. Everything that would substantiate the product thesis - ARR trajectory, agent activity, the Mercor headcount proof point, the accountant-shortage statistic - is company-supplied, single-sourced, and in one case described differently by the two publishers. No audit, benchmark, customer confirmation or incumbent response appears in the supplied material.
Real installed base, self-reported scale
Adoption is more than a pilot story: more than 600 customers is corroborated by both publishers, named logos span AI labs and public companies, roughly 40 percent are outside tech, and an EY alliance plus claimed partnerships with over half the Accounting Today top 20 CPA firms constitute genuine channel traction in a category that gates on accounting-firm endorsement. It is held below the upper band because every quantitative adoption datapoint originates with the company, the ARR metric is described inconsistently, no revenue level or retention is disclosed, and the single named deployment carrying the efficiency argument is unconfirmed.
Claims run ahead of shown evidence
The rhetoric - incumbent ERPs as pre-AI relics, the general ledger as the operating layer, an investor asserting customers run finance teams a tenth the traditional size, 'that vision is now reality' - rests on unaudited vendor metrics and a single unconfirmed customer example implying roughly $667mn of ARR per finance employee. The messaging also carries an unresolved contradiction between 'lean, impactful finance teams' and 'we're not coming after jobs'. The gap is positive but moderate rather than extreme, because the customer base, non-tech diversification and Big Four/CPA alliances are concrete and the financing is fully corroborated.
Vendor-controlled announcement cycle
The information chain is almost entirely interested. Rillet timed and controlled the disclosure, granting Fortune an exclusive; the growth metrics, customer proof point and workforce statistic come from the CEO's own statements and X posts; the lead investor, which just marked its own earlier position up and took a board seat, supplies the corroborating quotes to both outlets. Both publishers reproduce these figures with attribution but no independent verification, and one is aggregating other outlets' reporting.
Confident on the deal, thin on the thesis
Two independent publishers dated the same day agree on the financing, investors, customer count and the EY alliance, which anchors the assessment. Confidence is capped because the cluster has only two sources, both derive their substance from one vendor disclosure, they disagree on how the central growth metric is defined, and no operational, technical or third-party evidence exists in the supplied material to test the product claims.
invest
Tiny corp wants Etched's numbers. Jane Street led $700M at $21B without publishing any1 distinct publisher
invest
Spark's $22M bet that the agent framework layer can stay independent1 distinct publisher
product
Canva's $7.1bn markdown is an inference bill, not a mood swing1 distinct publisher
product
Gamma says it hit $100m ARR with 50 people, and 15 months of profit to go with it1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 19, 2026
1 article · August 19, 2026