Julian Rowlands’s road to the finance chief’s office began with a small statistical calamity: roughly one hundred job applications and one offer. The offer was not to forecast revenue, shepherd a fundraise, or explain margins to a board. It was to work with data systems for hedge funds. For four years, he learned what happens beneath the polished report - warehouses, labels, pipelines, and the stubborn fact that information rarely arrives in the shape its user would prefer.
This turned out to be an excellent education for a job he was not trying to get. Rowlands had studied history at Rutgers, then attended the University of Chicago Booth School of Business, where he teased the place for being a “CFO factory.” He wanted emerging markets. He went looking for businesses in difficult places, working across parts of Africa and Asia, including Uganda, Mali, Kenya, Bangladesh, and Myanmar. Finance leadership still seemed like somebody else’s career.
Then, while living in Yangon and doing startup investing work, he saw an opening at Xendit, an Indonesian payments company. It needed a CFO who could handle large amounts of data, speak the language of finance and venture capital, and live comfortably in Jakarta. Most conventional CFO candidates would have possessed more finance-team experience. Few would have matched that particular handful of requirements as neatly. Rowlands had never worked inside a finance organization. For this one peculiar job, he made sense.
An apprenticeship with the title already attached
Rowlands accepted the role and became Xendit’s first CFO in 2018. This is the sort of sentence that can be made to sound preordained after the fact. He tells it with more useful honesty. “I’m kind of an accidental CFO,” he has said. There was no secret certification hidden in a desk drawer. He told Xendit plainly that he had not done the job before and was not a CPA. The company saw someone who could learn and build.
His first move was to ask who could help him get smart quickly. Through his own contacts and Xendit’s Y Combinator network, he assembled what an interviewer later called a personal board of CFOs. Cam Hyzer, later the CFO who took ZoomInfo public, became a mentor. Other conversations included finance leaders from Segment, Brex, and OpenAI. Rowlands remembers walking through downtown San Francisco with Brad Lightcap, then OpenAI’s CFO, talking through what mattered, what could wait, how to form a team, and how to ready a company for scale.
“I’m kind of an accidental CFO.”Julian Rowlands
There is a fine distinction between confidence and theatre. Rowlands had enough confidence to take the job and little interest in pretending the title had supplied the experience. He worked long weeks, asked experienced people direct questions, and learned in public. The informal board did not do the work for him. It helped him recognize which work mattered.
Xendit supplied an accelerated course. When he arrived, the company had roughly 70 or 80 employees. Developers worked from the basement of a house in South Jakarta; there are videos, he says, of the basement flooding. By the time he left at the end of 2019, the team was around 250. During his tenure, he helped raise a Series B of more than $60 million from investors including Accel, Kleiner Perkins, and YC Growth.
The recurring report that kept recurring
Back in New York, Rowlands joined Spruce in February 2020. He was in the office for only a few weeks before the pandemic rearranged office life. The proptech company echoed Xendit in useful ways: a complex, transaction-heavy industry made simpler through an API. He started in FP&A and became head of finance as Spruce grew from about 75 people to roughly 250. In 2021, the company raised a $60 million Series C. Zillow later acquired it.
Across both companies, Rowlands kept meeting the same expensive nuisance. The information a CFO needs lives in accounting software, payroll systems, customer databases, budgets, and spreadsheets. The labels disagree. The structures disagree. A customer’s name may vary from one system to another. Before a finance team can explain performance, it must first assemble reality.
The bottleneck he chose to build around
Well-resourced organizations solve this with connectors, a data warehouse, business-intelligence software, and analysts who turn dashboards into a narrative. It works, but it can demand substantial time and money before anyone answers the original question. Rowlands had seen similar data problems in his first job. He had also lived them from the CFO’s chair. His résumé, once a collection of detours, had become an unusually precise product brief.
He started Cashboard in spring 2022 and entered Y Combinator’s Summer 2022 batch. The company’s promise was a single source of truth for finance: connect the scattered systems, map their labels, preserve the definitions of important metrics, control permissions, and automate recurring reports. Its North Star earned a compact phrase: “inventing the solo CFO.”
Give the machine a dictionary
The rise of generative AI sharpened rather than erased Cashboard’s original argument. A model can write an analysis quickly. Finance cannot tolerate a fresh interpretation of gross margin every month. If the underlying mappings drift, the prose may become smoother while the business meaning becomes less reliable. Rowlands’s answer is a governed semantic layer: live data, approved mappings, canonical metrics, and access rules that travel together.
It is a conspicuously unromantic foundation for an AI company. Dictionaries and permissions rarely enjoy keynote lighting. Yet the dull parts are where trust lives. Cashboard connects finance data and makes that context available through dashboards, Excel, and AI tools. The company says its platform supports more than 900 integrations. In July 2026 it announced a $5 million seed round led by FINTOP, with TTV Capital participating, bringing disclosed funding to $6.9 million. The capital is intended for go-to-market expansion, product development, and an agentic budgeting and forecasting product.
“The North Star of the business is inventing the solo CFO.”Julian Rowlands
“Solo” can sound like a head-count claim, but Rowlands frames it as an aspiration about leverage. The CFO should spend time making decisions and explaining the business, while software handles repeatable assembly and reporting. Earlier in his career, he hired an FP&A director at Spruce when the meeting load outgrew his calendar. He is not romantic about one person doing everything. He is interested in how much of the preparatory work a reliable system can absorb.
His public advice to finance leaders experimenting with AI is characteristically practical: “Take a lot of demos.” Use the tools, learn their limits, and do not confuse a disappointing spreadsheet prompt with the ceiling of a purpose-built product. He draws a line between tasks suited to language models and calculations better handled deterministically. Enthusiasm, in this version, comes with a calculator and a reconciliation check.
The advantage hiding in the detours
Rowlands has described himself as creative, scrappy, and hard-working. His short public biography adds “scruffy & kind,” calls him a Jersey kid, and notes that he is the son of a writer. The combination explains some of his appeal as a finance founder. He can speak in systems without sounding like a manual. In one podcast exchange, after the host joked about eliminating VLOOKUP and asked about INDEX MATCH, Rowlands replied: “I’m an XLOOKUP girl.” It is difficult to make spreadsheet syntax charming. He found a way.
He also writes beyond product announcements. In 2021, while at Spruce, he published an essay arguing that “Distributor APIs” deserved recognition as a distinct category. These companies sit between old industries and software customers, distributing a commodity service through an API and earning money through transactions. The piece grew from seeing conventional SaaS metrics applied where they did not quite fit. Again, the instinct was to notice mismatched labels and supply a better definition.
The most reusable idea in his story has little to do with finance. Rowlands advises would-be founders to know what they are particularly equipped to build. He does not claim an all-purpose founder identity. His fit is narrow: startup finance, data infrastructure, complex operating environments, and a tolerance for uncertainty. Cashboard occupies that intersection.
Careers are often cleaned up for publication until every step appears to anticipate the next. Rowlands’s version is more generous. One offer led to data systems. A wish to work in emerging markets led to Yangon. A job listing led to Jakarta. Admitting what he did not know led to mentors. Two finance roles led to the same recurring frustration, and the frustration led to a company. The route was crooked. The accumulated knowledge was not.
A history major may be especially qualified to recognize that the meaning of an event changes with the context around it. A strange résumé is strange only until the right problem arrives. Then the detours become the map.