San Francisco's quiet fix for one of the office's loudest paperwork problems - parental, medical, caregiver, and disability leave - now serving more than a thousand employers.
Filed Tuesday, San Francisco, CA - the office where HR finally stops crying.
It's 9:14 a.m. somewhere in a 600-person company. A payroll lead opens her laptop, sees three new parental leave requests in her queue, and does not flinch. A year ago, this was the moment her week ended.
Sparrow sells software that no one wants to need and everyone eventually does. The company runs employee leave - parental, medical, caregiver, ADA, anything that pulls a worker out of their seat for more than a few days - as a managed service for employers. It is unglamorous work. It is also, by most accounts, the worst job in HR. Filings in fifty states. Insurance carriers that fax. Disability paperwork that asks the same question in three places and rejects you if the answers do not match. Sparrow's bet is that you can automate most of this, and that the parts you cannot automate are precisely where a human being should show up. They call it high-tech, high-touch. Their customers, including OpenAI, Reddit, and Oura, just call it the reason they sleep.
Leave is a moment, not a ticket.- the Sparrow product principle, paraphrased
The HR industry has, for two decades, been busy building beautiful software for the parts of work that already function. There are dashboards for hiring. Dashboards for performance reviews. Dashboards, even, for dashboards. Employee leave - the most emotionally loaded, legally complex, and time-sensitive workflow any people team ever touches - was largely left to spreadsheets and goodwill.
The reason is simple, if uncharitable. Leave is rare per employee but devastating per HR rep. A people team of three at a 500-person company will handle maybe forty leaves a year. Each one consumes between twenty and forty hours of work: gathering doctor's notes, filing state forms, coordinating short-term disability carriers, syncing payroll, answering panicked Slack messages from a new parent at 11 p.m. The math is brutal and the software market, for a long time, decided the math was somebody else's problem.
It also happened to be exactly the kind of work that breaks at scale. A startup grows from sixty employees to six hundred and the head of people discovers, sometime during the third concurrent parental leave that month, that the system is held together with sticky notes and one very tired person named Janet.
The most human moments at work were the ones the software treated worst.- the founding observation, more or less
Deborah Hanus is, by any reasonable accounting, overqualified for paperwork. She picked up multiple engineering degrees at MIT, did reinforcement-learning research at Google Brain, taught computer science to monks in Cambodia on a Fulbright, and started a PhD in artificial intelligence at Harvard. Then, in 2018, six of her close friends took leave from work at roughly the same time. All six had a miserable experience. None of them were sure how to file. Some of them got paid weeks late. One did not get paid at all.
She left the PhD. With co-founder Marek Sotak, she started Sparrow with a thesis that sounded almost suspiciously simple: nobody was going to fix this by hiring more HR people. The fix was software that could swallow the compliance complexity, paired with humans whose entire job was showing up for the employee. Half engineering team. Half leave specialists. One product.
Investors, to their credit, eventually got it. Initialized Capital led an early round. GGV came in for the Series A. By the summer of 2025, SLW led a $35 million Series B that brought Sparrow's total raised past $55 million. The bet, by then, was working.
The Sparrow product is, mechanically, a leave-management platform with three layers stacked on top of each other.
The first layer is intake. An employee in California announces a baby on the way, or a back surgery, or a parent who needs hospice care. They tell Sparrow what's coming. Sparrow's software figures out which combination of federal, state, and local leave laws apply, what paid leave the employee is entitled to under their employer's policy and the relevant jurisdictions, and exactly what needs to be filed where.
The second layer is automation. Sparrow files. It coordinates with state agencies, short-term disability carriers, and the employer's HRIS. It pushes status updates into Slack. It rewrites a payroll plan every time a leave gets extended, shortened, or moved.
The third layer, and the part that quietly closes deals, is the human one. Every leave gets a Sparrow leave specialist. They answer the 11 p.m. Slack message. They explain why the state denied a claim and how to refile. They are the reason HR teams who try Sparrow tend not to leave.
The software does the paperwork. The humans do the part where it matters that someone listened.- how Sparrow describes itself, with the marketing varnish off
It is one thing to claim that a leave-management product saves HR teams time. It is another to put numbers next to the claim and let readers do their own arithmetic.
OpenAI runs its leaves through Sparrow. So do Reddit and Oura. The reason these companies, all of whom could afford to build a leave-coordination team internally, choose not to, is the same reason most companies stopped running their own mail rooms. It is not core to the business and it is harder than it looks.
You can tell a company cares by how it handles the worst week of an employee's year.- a sentiment Sparrow's marketing keeps returning to
Sparrow's stated mission is to make employee leave stress-free, for employers and employees both. That sounds like a tagline. It also happens to be the kind of operational mission that survives contact with reality, because the company can prove it: hours saved, claims approved, payroll plans rewritten on the fly when a leave extends by two weeks.
What Sparrow is not is a benefits brokerage. It does not sell insurance. It does not replace the carrier. It sits in the workflow between the employer, the employee, the state, and the carrier - the place that, before Sparrow, was nobody's job and everybody's problem.
Paid leave law in the United States is, charitably, a patchwork. Eleven states plus D.C. now run their own paid family-and-medical leave programs. More are coming. Each new program adds a new form, a new portal, a new set of rules about who is eligible and how to file. The complexity is not going to shrink. Someone has to absorb it.
The simpler reading is that Sparrow is a back-office utility for HR teams. The longer one is that it is a small piece of social infrastructure - the kind of thing that decides whether a new parent gets their paycheck on time, whether a person on chemotherapy can stop worrying about whether their job will still exist in eight weeks, whether a caregiver can take three months off without bankrupting themselves on the way out the door.
That is the work. It is not glamorous and the founders, refreshingly, do not pretend otherwise.