Nami Baral

The through-line

Nami Baral's recurring argument is that the biggest companies get built in the markets everyone else has written off, and only by people willing to spend years underground before anything shows above the surface. She chose payroll in 2022 precisely because it looked dull: "sleepy markets are interesting because they give you a contrarian, you know, way to approach where value can be built" [1]. That contrarianism runs through every decision she describes, from taking two and a half years to build before going to market, to owning the payroll tax engine rather than renting someone else's, to building AI agents at a moment when almost nobody was [1]. Underneath it is a personal method she has applied since leaving Nepal for college over a dial-up connection: form a view on where the world is going, then position yourself there. "Opportunities don't just come to you. You have to actively be, you know, in a position where you increase the surface area for lightning to strike" [1].

The second thread is time horizon. She frames her work in decades rather than quarters, and rejects the metric most founders optimise for. "What is the fastest path to 100 million in ARR? That is not really just the question I'm thinking about" [1]. Her stated ambition is a category that does not exist yet, and she is explicit that the price of that ambition is patience, both hers and her investors' [1].

On why payroll is the right sleepy market

The insight that drew her in was fragmentation across a company's lifecycle. Companies start with one payroll vendor, migrate to another, then another, and simultaneously run several vendors for the same workforce: "you think of your team as one, but why do you have to use Deel for international, ADP for domestic, benefits with somebody else, your vendor payments with somebody else, it just does not make sense" [1]. Her answer is a single comprehensive system of record covering EOR, PEO, contractor payments and international local entity payroll, capabilities she describes as spanning what ADP, Workday, Gusto, Rippling and Deel each do separately [1]. The market appeal is total addressability: "This is one of those categories where the TAM is endless. Every company in the world could be a Neural client" [1]. She wanted a product that works for a company "from day one to IPO and beyond" [1], and describes the platform as all-in-one payroll, team management and compliance for global hiring [9], spanning payroll, HR, benefits and payments as businesses go global [3].

The timing argument matters as much as the market argument. She is clear that this was not buildable before: attempting the same consolidated stack ten years ago "would have taken me 20 years to build" without the recent technological innovation [1]. AI agents are what make a single end-to-end system of record achievable in a reasonable window.

On building for years before going to market

She took two and a half years of heads-down building with no go-to-market, and defends it on first principles: "to build a truly generational company, you cannot take shortcuts" [1]. The doubt was real and constant, especially reading the news as a founder, but she resolves it by scale of horizon. She expects to spend decades on this, so "1 or 2 years is just a blip for me for the amount of time I'm going to really be building this" [1].

Her metaphor for this is the Chinese bamboo, which shows no visible growth above ground for about four years while its roots spread, then grows ninety feet in roughly four weeks [1]. The point is not patience for its own sake but what the roots make possible later. She contrasts her approach with companies that "build for payroll just build for payroll," take the easy route with someone else's tax engine and an FBO structure at a local bank, and never lay "the concrete foundation for all kinds of money movement, which comes with its own set of regulatory" requirements, across 150 countries [1]. She acknowledges the current environment produces companies reaching revenue in record time and treats that as part of the same journey, while insisting the foundation had to come first [1].

On the ambition beyond payroll

Payroll is a wedge, not the destination. Her framing: payroll is one form of money movement, often the largest scheduled outflow a company makes, but only one, and "the same structure that can move payroll for related funds should be able to handle every inflow, every outflow, and every financial decision that the CFO suite makes" [1]. The sequencing is deliberate: "solve that for payroll first, earn the right to then be able to get the trust to move other forms of money for that company," expanding from payroll into contractor payments, vendor payments and accounts receivable [1]. She sells to the CFO, not only the head of HR [1].

She states the end state bluntly: "In my mind, it's ADP, Oracle, and Visa combined, right? And that's a category that is yet to exist because laying the foundation for something like that is so difficult to do" [1]. Hence the deliberately outsized revenue target she raises against the usual hundred-million benchmark, asking instead what the fastest path to a hundred billion looks like and "what is the infrastructure that you put in today" [1]. She grounds the confidence in comparison: hundreds of payroll companies with billions in revenue each solve point problems, while her company already covers the full set as a roughly three-and-a-half to four-year-old business, which she calls an engineering feat in itself [1].

On AI at the DNA rather than bolted on

She draws a hard line between AI-native architecture and surface-level AI, describing agents running on "each and every module of the company. From fraud detection to transaction monitoring to KYB to payroll processing to, you know, international state compliance," rather than "AI that is bolted on as if it's a ChatGPT wrapper" [1]. She was building AI agents from 2022, before the ChatGPT wave arrived and when doing so was genuinely difficult [1], and points out that her previous company, Harvest, was "the original AI agent in consumer fintech, way before anybody was building AI agents" [1][6]. The commercial consequence is efficiency: a burn ratio under 0.5 with roughly 120 to 125 people on the team, because "the tech is over working overtime for us" [1]. She has also spoken more broadly about how AI is reshaping finance and global business, and about preparing for a future that includes AI workers [3][6].

On what technology cannot solve

Her counterweight to the AI story is that some moats have to be negotiated rather than engineered. Benefits are the example: "You cannot run a proper company payroll without giving benefits to your employees" [1]. She describes going live with Aetna on a special master plan relationship, going live with Guardian, and going live with Kaiser and MetLife, framing these institutional carriers as lending credibility to a newcomer [1]. Her reading of why this matters is commercial rather than sentimental: bringing large institutions along for the journey "is what creates stickiness" [1]. A generational company, in her account, is a combination of the technology and those relationships, not either alone.

On picking the people and the capital

Conviction is only half her stated secret sauce; the other half is who you bring. She handpicked investors "patient enough to see and wait for that next moment to arrive," describing them as patient capital that knows where the puck is going [1], and treats knowing which investors would buy the vision as a privilege of being a repeat founder. On team, she went back to people she had already pressure tested: previous co-founders and previous team members, and credits "the power of that network" with building the product in record time [1]. The profile she wants is people who will put in 200 percent, "extremely loyal, extremely low ego, and heads down the best builders" [1]. For her second company she deliberately brought on multiple co-founders so there would be meaningful help while building [9]. She has also discussed the mechanics of fundraising itself, including raising $7 million in two weeks [7], and approaches investing from the other side of the table as an angel [6].

Culture, for her, includes moving opportunity across borders: taking the best employees from the Nepal office and bringing them to the US, giving them the same kind of opportunity she had, which is also part of how the company operates so capital efficiently [1]. She has talked elsewhere about meaningful work as something that comes from a sense of community and alignment with personal values [4], and about being a woman in technology across three startups [2], including doing it as the mother of a young toddler [5].

On calculated risk and forming a viewpoint

Her self-description is precise: "the best founders tend to be calculated risk takers. I'm not somebody who will do a lot of risk taking in, let's say, adventure tourism, but I will take a lot of really interesting risk when I feel like, you know, uh the world is moving to that place" [1]. The origin story she offers is applying to US colleges from Nepal between midnight and 4am on a dial-up connection where a single page took three minutes to load, reasoning that "the more information I have, the more opportunities I have" [1]. The same logic carried her out of investment banking into a zero-revenue startup that was acquired by Twitter pre-IPO, then out of Twitter in 2018 to found Harvest, sold to Acorns, and then into payroll [1][6][7]. She describes these as calculated bets taken all her life that built the perspective and confidence to launch her own companies [9], and at Harvest specifically, iterating toward product-market fit and going all in once the aha moment arrived [9]. The common theme she names across all of it is having a viewpoint on where the world is going, which is something you have to actively work at rather than wait for [1].

Takeaways

• Choose sleepy markets deliberately: they offer a contrarian route to value that crowded categories do not [1].
• Fragmentation is the opening in payroll, where one workforce is split across Deel for international, ADP for domestic, and separate vendors for benefits and payments [1].
• Two and a half years of building before go-to-market is justified by horizon, not speed: "to build a truly generational company, you cannot take shortcuts" [1].
• Payroll is a trust wedge into all CFO money movement, earned by solving payroll first before contractor payments, vendor payments and receivables [1].
• The stated end state is a category that does not yet exist, described as ADP, Oracle and Visa combined [1].
• AI-native means agents on every module, from fraud detection and KYB to compliance, which is what sustains a burn ratio under 0.5 at roughly 125 people [1].
• Technology alone does not create stickiness: direct master plan relationships with carriers including Aetna, Guardian, Kaiser and MetLife supply credibility a young company cannot code [1].
• Hire people you have already pressure tested, and select investors for patience rather than availability [1].