The bookkeeping company that thinks like a software company
It is the last week of the quarter, and somewhere in a Series B startup's Slack a founder is not panicking about the close. That is the Pilot trick. The numbers tick into place inside a dashboard. A real accountant - American, salaried, named - has already flagged the one weird transaction. The investor update writes itself, more or less. Bookkeeping, the most boring sentence in business, becomes a sentence the founder doesn't have to read.
Pilot.com sells finance as infrastructure. Bookkeeping, taxes, fractional CFO work - the things every founder hates and every investor demands - bundled into a monthly subscription and run on Pilot's own internal software. The company headquarters is in San Francisco at 180 Montgomery. The headcount is around 300. The customer list spans SaaS to e-commerce to biotech, and includes, depending on the year, a meaningful share of the venture-backed market.
Bookkeeping is boring. Getting it wrong is catastrophic.
For decades the choice for a young company was a familiar one: a part-time bookkeeper who used QuickBooks the way you use a TV remote you can't fully remember, or an accounting firm whose bills arrived monthly and whose answers arrived eventually. Either way, the founder ended up in the spreadsheet anyway. Not because they wanted to. Because the numbers were too important to trust.
This is the problem Pilot's founders kept running into. They ran two startups before this one - Ksplice, sold to Oracle, and Zulip, sold to Dropbox - and at both companies the back office was a tax on the founders' attention. The third time they sat down to start something, they decided the back office was the company.
An origin story buried in a finance review
When Oracle bought Ksplice, the finance team noticed that the founders had built internal bookkeeping tools that were, frankly, better than the ones a $200B company was using. Somebody said so out loud. The founders filed the comment away. It came back, years later, as Pilot.
Three friends from MIT, betting a third time
Waseem Daher, Jessica McKellar, and Jeff Arnold met as undergraduates in MIT's student computer club. They are, by their own description, three computer scientists who could have built almost anything. They have, in fact, built three companies together - and, more impressively for any three co-founders, still answer each other's emails.
Their bet on Pilot was a strange one. Bookkeeping is a market famously hostile to software-only solutions. Customers want a name, a person, an inbox that replies. The founders' insight was a humbler version of their MIT instincts: build the software, but staff the company with accountants too. The product is not the dashboard. The product is the quiet feeling that the books are done.
A $400,000 domain name
Shortly after raising their seed in 2017, the team did something most early-stage founders would call reckless: they spent $400,000 to buy the pilot.com domain. The company had been called Zapgram. Two months later it wasn't. The domain, in retrospect, was one of the cheaper line items on the balance sheet.
People, plus software, plus discipline
Pilot's offering breaks into three layers, sold separately or together: monthly bookkeeping, tax preparation (including the R&D credit work most founders never claim because nobody told them to), and a fractional CFO service that gets called in when the deck needs forecasts and the forecasts need to be defensible. The pricing scales with expense volume. The accountants are American and named. The software is invisible until you want it.
What sets Pilot apart is less the feature list than the boringness of the promise. Other vendors sell you accounting software and let you find a human. Pilot sells you the human and lets them find the software. For a busy founder, that is a meaningfully different sales pitch.
A short timeline of a quiet company
Founded as Zapgram. Renamed Pilot after $400K domain purchase. Seed round, led by Index Ventures.
$15M Series A, also led by Index.
$40M Series B. Stripe joins the cap table - a notable customer becomes investor.
$60M Series B-1 in February; $100M Series C in March led by Bezos Expeditions. Valuation: $1.2B.
Two rounds of layoffs amid the tech downturn. Pilot, like its customers, learns to operate leaner.
Continued focus on tax credit work, CFO services, and integrations with the modern finance stack.
Sequoia, Index, Bezos - and the math behind the cap table
By the time Jeff Bezos's personal investment vehicle led Pilot's 2021 Series C, the company had a track record few bookkeeping startups achieve: real revenue, low churn, and a customer base that pitched the product to their own portfolio companies. The $100M round was not a bet on growth at any cost. It was a bet that finance, properly tooled, becomes the kind of recurring software business venture investors recognize.
Pilot.com funding history
Pilot integrates with the now-standard founder stack: Stripe, BILL, Gusto, Rippling, Deel, Carta, plus the usual cloud accounting backbones. The integrations matter less for marketing reasons than for plumbing reasons. The fewer places a customer has to copy a number, the fewer places that number goes wrong.
Make the back office boring on purpose
The stated mission is unromantic, which is the point. Take the back-office work that drains founders' time, and turn it into reliable, software-driven infrastructure. The unstated mission is bolder: prove that managed services - real humans doing real work - can scale like SaaS when the software is good enough. If Pilot is right, a generation of operationally heavy businesses get re-imagined the same way.
The Pilot card
- FOUNDED: January 2017, San Francisco
- FOUNDERS: Waseem Daher, Jessica McKellar, Jeff Arnold
- HQ: 180 Montgomery Street, San Francisco, CA 94104
- EMPLOYEES: ~300
- BACKERS: Sequoia, Index, Bezos Expeditions, Stripe, Whale Rock
- VALUATION: $1.2B (as of March 2021)
- NAICS / SIC: 54121 / 8721
The back office becomes software, finally
If you squint, Pilot is a bet about the next decade of work. Generative AI is now eating the parts of bookkeeping a human used to do by hand - categorizing transactions, reconciling statements, drafting management reports. The companies positioned to win are the ones that already pair software with experienced humans, because their humans can move up the value chain instead of being replaced by it. Pilot has been quietly building that org chart since 2017.
The skeptic's case is also fair. Managed services have unpleasant unit economics when growth slows, which is why 2023 brought layoffs and why 2024 brought more. The optimist's case is the one Bezos wrote a check for: the market for trustworthy financial back-office is enormous, the moat is the trust, and the trust compounds the longer a customer stays. Pilot is, structurally, a business built to keep its customers a long time.
So back to the founder, last week of the quarter, not panicking. The Slack channel is calm. The investor update is half-drafted in a Google Doc. The numbers were ready before the question was asked. There is no version of this scene where the founder thanks Pilot - that is also the entire point. The companies that build infrastructure properly tend to disappear into the wall. Plumbing, after all, is most useful when nobody is talking about it.