Month-end has a peculiar power over adult calendars. Dinners become provisional. Weekends acquire footnotes. A company may sell software in real time, pay people in real time and watch customers arrive in real time, yet its clearest view of itself appears weeks later, after accountants have reconciled the evidence. Nicolas Kopp lived inside that lag while running N26’s American business. His finance team, he says, was excellent. The wait was not. Basic metrics could take four to six weeks to become reliable, at which point the business had already made several new decisions.
Most people respond to slow reporting by requesting another dashboard. Kopp responded by founding an enterprise resource planning company, which is the corporate equivalent of noticing a loose floorboard and deciding to rebuild the house. Rillet, the company he started with engineer Stelios Modes in 2021, goes after the general ledger itself: the central accounting record where a business finally decides what happened.
It is difficult software to make and even harder software to sell. A cheerful new note-taking app may ask for your thoughts. A new ERP asks for the books. Kopp understood the distinction. Rillet spent three years building before the wider world saw the product. For roughly two and a half of those years, it stayed in stealth. Its first customer was itself.
“Building any company is hard, you may as well try to go for something a bit harder.”Nicolas Kopp, on choosing the mission
An education in financial plumbing
Kopp’s career has the tidy progression of someone following money deeper into the machine. The Swiss founder studied business administration at the University of St. Gallen, then accounting and finance at the London School of Economics. He spent five years at Morgan Stanley in London and Hong Kong. The work supplied fluency in markets and financial statements. It also produced an important connection: during a Morgan Stanley assessment center, he met Valentin Stalf, who would later co-found N26.
In 2015, Kopp joined N26 when the Berlin neobank had only a few dozen employees. He has recalled signing his contracts in what was effectively a small utility room. The setting suited the job. As head of business development and operations, he became a generalist, building the function, helping establish partnerships such as TransferWise, working on premium accounts and leading early international expansion. In 2017, N26 sent him to New York to build the US business.
That period gave Kopp an operator’s view of regulated infrastructure. N26 was not merely drawing a nicer checking account on a phone. The team was building partnerships, compliance processes and payment rails while expanding across borders. Kopp sat on the business and finance side; Modes, then an infrastructure architect at N26, worked on payment systems. Years later, their skills would meet again around a less visible but equally consequential system.
N26’s American expansion also supplied a less comfortable lesson: ambition does not exempt a company from the particulars of a market. The neobank eventually withdrew from the United States. Kopp does not present that chapter as a neat victory lap. What endured was the operator’s reflex to inspect dependencies, learn a customer’s workflow and respect the unglamorous systems that decide whether a product can function. Rillet would need all three. Accounting software is full of grand promises that eventually meet a chart of accounts.
The ledger is late to its own meeting
Traditional ERP systems are systems of record. Transactions arrive, people classify them, approvals move through workflows, and the result is stored. Much of the actual reasoning still happens elsewhere: in spreadsheets, messages and the heads of experienced accountants. At the end of the month, teams gather the scattered story and persuade it to balance.
Rillet’s argument is architectural. If source data from banks, billing platforms, payroll tools and customer systems arrives continuously, the accounting work can also happen continuously. Software can match transactions, prepare schedules, flag exceptions and draft entries as events occur. Humans retain approval authority and a traceable record of why each decision was made. The goal is not an unsupervised robot with a company credit card. It is a ledger that arrives prepared for the conversation.
From business event to explainable books
Kopp calls the destination a “system of context.” A record should contain not only the transaction, but the policy, approval and evidence that explain it. This is the useful dividing line between an AI party trick and financial infrastructure. In accounting, confidence without provenance is merely a faster route to a difficult audit.
His own habits reflect that respect for detail. Kopp has said he continues to do work inside Rillet even when observers suggest it is not a chief executive’s best use of time. His answer is emphatic: using the product is part of the job. He advises finance buyers to know vendors closely and test whether they understand the actual workflow. Occasionally, he asks a prospect why a particularly sharp question came up. Some admit they were testing him. He seems to enjoy the exam.
“Do the right thing. Care about your customer. Build an awesome product. Improve that product every day.”Kopp’s compact operating doctrine
Trust first, velocity second
When Rillet launched publicly in August 2024, it was about ten people asking companies to place serious records in unfamiliar hands. By the end of that year, the company said it had 100 customers. In May 2025, Sequoia led a $25 million Series A. Roughly ten weeks later, Andreessen Horowitz and ICONIQ co-led a $70 million Series B. The company reported more than 200 customers at the time.
Then came a much faster chapter. In August 2026, after a board meeting in which the company shared its growth, Rillet raised $100 million in a round Kopp said came together in less than 48 hours. ICONIQ led the Series C, with returning support from Sequoia and Andreessen Horowitz and participation from other investors. The financing valued Rillet at $1 billion and took reported total capital above $200 million.
The speed of the round makes a fine headline, but the slow part is more revealing. Enterprise accounting is won by accumulated evidence: a successful implementation, a reconciled account, a clean audit trail, a controller willing to recommend the system to another controller. Kopp’s repeated emphasis on customer service is less decorative than it sounds. In a trust product, support is part of the architecture.
The investor list traces the widening circle around the idea. Early backing came from First Round Capital, Creandum, Susa Ventures and BoxGroup, alongside finance operators who understood the buyer. Later rounds brought Sequoia, Andreessen Horowitz, ICONIQ and Oak HC/FT into closer roles. Board seats followed. Kopp’s most revealing answer about a future moat, however, was not a patent or a proprietary model. It was velocity and customer care. In enterprise software, those two virtues are often in tension. Moving quickly is only useful if the books remain boringly correct.
Rillet now says it supports more than 600 customers, including public companies and enterprises with billions in revenue. It also says customers are moving from established systems such as Oracle Fusion, SAP, Workday and NetSuite. Those claims mark a shift in the question facing the company. Early on, the puzzle was whether anyone would entrust a ledger to a startup. Now the puzzle is whether the culture that earned those first accounts can survive hundreds more.
A more interesting accountant
Kopp’s aspiration reaches beyond a shorter close. He argues that accountants and controllers understand the business more intimately than almost anyone, yet routine reconciliation consumes much of their time. If software handles more matching and preparation, their judgment can move toward policy, analysis and decisions. The profession does not become less important. Its least imaginative chores become less dominant.
There is a humane observation inside the automation pitch. A spreadsheet is not offended by a canceled dinner, but a person is. Close week has long converted system limitations into personal sacrifice. Rillet wants the work to occur as the business happens, leaving fewer surprises to be discovered in a monthly siege.
Of course, every established software company can add an assistant. Kopp’s answer is that a native system begins with a different assumption. Old software treats people as the engine and optimizes their forms. Rillet treats financial events as material to reason about, then asks people to supervise exceptions, approve judgments and examine proof. Whether that architectural distinction becomes a durable advantage will be tested in implementations, not slogans.
Kopp appears to like that sort of test. He chose accounting partly because it was intellectually difficult and consequential. Before ERP became a lively venture category, the probability-weighted case for starting Rillet looked rather rude. He did it anyway, with a co-founder who knew the plumbing and a team that included accountants, auditors and finance operators.
The bet is now large enough to attract attention, money and imitation. Yet its originating scene remains modest: a chief executive waiting weeks for a number, despite having good people and digital data. Rillet’s future rests on whether it can make that wait feel as antiquated as balancing a ledger with a quill. There are easier ways to build a software company. Few are so neatly reconciled with the founder’s past.