The San Francisco startup pairs AI agents with humans who stay awake in shifts, handling the hiring and scheduling that keeps home care agencies from growing. It started with a caregiver who didn't show up.
On her first day of home care, Nick Walker's grandmother waited for a caregiver who never came. Nobody had a good explanation. The agency was short-staffed, the schedule had a hole in it, and a phone line somewhere went unanswered. It is the kind of small failure that rarely makes news and happens constantly, and it is the reason MochaCare exists.
MochaCare, part of Y Combinator's Winter 2026 batch, describes itself plainly: an agentic management service for care organizations. In practice that means it takes over the parts of running a home care agency that owners hate and cannot escape - hiring caregivers, filling shifts, handling client intake - and runs them with AI agents that are watched, around the clock, by humans. The pitch is not that software will do the caring. It is that software plus a few very awake people can keep an agency from dropping the ball.
That distinction matters more than it sounds. Most startups aimed at healthcare operations sell a dashboard and wish the customer luck. MochaCare's two founders decided the dashboard was the problem. Agency owners do not have time to learn another tool. They have time to hand a job to someone who will actually do it.
Home care - the business of sending caregivers into people's homes to help aging or disabled clients with daily life - is enormous and strangely under-served. It is a market worth well over $130 billion a year in the US, and it grows every year as the population ages. Yet the agencies that deliver it are often run on a patchwork of spreadsheets, text messages, and whatever legacy software they signed up for a decade ago.
The single hardest number in the business is turnover. Caregivers leave at rates north of 70% a year, which means an agency owner is never really done hiring. The moment a pipeline of new caregivers slows, shifts start going uncovered, clients get let down, and referrals dry up. Growth in home care is less about clever marketing and more about whether you can answer the phone at 11pm when a shift falls through and find someone to cover it.
"Real humans + AI tools, awake 24/7 to support you."MochaCare's tagline
This is the gap MochaCare is walking into. Not a flashy consumer app, not a research lab moonshot, but the operational grind that keeps a very large and very human industry from working the way it should.
It helps to picture the person on the other end of the product. A typical home care agency is not a hospital system with an IT department. It is an owner, often a former nurse or caregiver, running a business out of a small office with a couple of coordinators. Their day is a stream of interruptions: a caregiver who overslept, a family asking for a schedule change, a stack of onboarding documents waiting on a signature, a new client referral that needs a call back before a competitor gets to it first. Every one of those tasks is winnable on its own. Together, they are why the owner never gets to the work that would actually grow the business.
MochaCare's product has two centers of gravity. The first is hiring. The company built what it calls the first AI-powered applicant tracking system made specifically for home care. Agencies configure their hiring pipeline once - the stages, the documents a caregiver has to submit, the checks that have to clear - and AI agents move candidates through each step on autopilot. An AI recruiter can run first-round interviews and coordinate the back-and-forth that usually eats an office manager's week.
The second is scheduling and coverage. MochaCare handles shift management, sends caregivers reminders, automates clock-in and clock-out and session notes, and takes on the after-hours scrambling that no one wants. When a shift is at risk, the system is designed to notice and act before it becomes a client-facing failure.
FIG. 1 - The handoff that defines the company: agents do the volume, a human stays on the feed for the moment it matters. The last box is the whole business model, painted navy.
Underneath both is a data layer. Because MochaCare sits across hiring and scheduling, it can see where an agency is leaking time and where it could grow - which referral sources actually convert, which shifts keep going uncovered, where a caregiver is about to churn. The company frames this as growth insight, and it is the natural payoff of doing the operational work rather than just storing the records of it.
Crucially, MochaCare does not ask agencies to abandon the tools they already run on. It integrates with the systems of record home care already uses, so it slots into a workflow rather than replacing it.
MochaCare was founded in 2026 by Nick Walker and Pranav Uppiliappan. Both were raised by their grandparents, which is not a line on a pitch deck so much as the reason the company has the shape it does. They came to the problem as people who had cared for elders, then built the software second.
The most-repeated detail about the pair is their schedule. To promise agencies genuine 24-hour coverage with a team of two, Nick works until roughly 3:30am and Pranav picks up from there. It is not a sustainable long-term staffing plan and they know it, but it is an honest signal of what the product sells: someone is always watching.
Both founders were raised by their grandparents. The company is what happens when engineers who have actually done the caregiving decide to fix the office instead of the bedside.
The strategic bet underneath MochaCare is that home care does not want more software - it wants the work done. So rather than selling a self-serve SaaS seat and charging by the user, MochaCare positions itself as a managed service. An agency hands over hiring, scheduling, or both, and MochaCare runs it as an ongoing operation, priced accordingly.
This is a deliberately harder business than shipping a tool. It means MochaCare is accountable for outcomes, not just uptime. But it also means the company is measured on something an agency owner feels immediately: fewer no-shows, faster hires, and evenings that belong to them again. The AI keeps the model economical; the human escalation keeps it trustworthy in a field where a missed shift is a person, not a support ticket.
There is also a defensibility argument buried in the approach. A pure software product in this space competes on features and gets compared line by line against the incumbents. A managed service that quietly runs an agency's operations becomes something harder to switch away from - it accumulates context about how each agency hires, who its reliable caregivers are, and how its clients like to be handled. The longer MochaCare runs an agency, the more it knows, and the more painful it becomes to hand that knowledge back to a spreadsheet.
FIG. 2 - Where an agency owner's week goes, and the slice MochaCare aims to absorb (illustrative).
The pitch in one chart: shrink the top three bars, hand the owner back the bottom one.
MochaCare's competition is less a rival startup and more the status quo: owners and office staff doing all of this by hand, plus the legacy platforms - WellSky, AxisCare, AlayaCare - that store the data but leave the labor to humans. Instead of trying to rip those systems out, MochaCare plugs into them and does the labor. That is a quieter go-to-market than a rip-and-replace, and probably a smarter one in an industry that changes software reluctantly.
The timing is not incidental. AI agents only recently got good enough to run a hiring pipeline or a scheduling queue with acceptable reliability, and the cost of keeping a human in the loop has come down enough to make 24-hour coverage affordable for a two-person team. A version of MochaCare attempted three years ago would have been all human and no leverage, or all software and no trust. The company is a product of the specific moment when those two curves crossed.
It is early. The team is two people and the company has raised a small first round tied to its YC batch. There are no splashy customer numbers to cite yet, and the honest read is that MochaCare is at the stage where the story is the founders, the market, and the wedge. But the wedge is real: a huge, aging-driven market, a brutal turnover problem, and almost no software built for the people actually running it.
The name is a small tell. Mocha - warmth and a jolt of caffeine - for a service that stays up all night so an agency owner does not have to. Whether MochaCare grows into the operating layer for home care or stays a promising W26 experiment will depend on how well those AI agents hold up when the human minding the feed is asleep. For now, the company is betting that in care, the answer is to make sure someone always is.