SafeinHome

A support company where the staff aren't in the room

Here is a business model that sounds, at first, like a contradiction: SafeinHome provides around-the-clock support to people with disabilities, and its support workers are almost never physically present. The company, founded in 2012 by Tim Lloyd and Lauren Yazolino and now run by CEO Ken Traverso, has spent more than a decade arguing that this is a feature. If you can deliver the reminder, the check-in, the reassuring voice at 3am without a caregiver standing in the kitchen, you get something valuable to two parties who rarely agree - the person, who gets privacy and autonomy, and the payer, who gets a bill that is roughly 60% smaller per support hour.

The mechanics are deliberately unglamorous. A wireless network of small sensors goes around the home and quietly tracks movement, activity and routine. A two-way tablet sits nearby, dark until someone taps it, at which point a trained staffer answers. The sensor data flows to what SafeinHome calls a "Care Circle" - family, providers, the support network - so the people who care about someone can know they're okay without hovering. It is, in a sense, the opposite of surveillance. The distinguishing fact isn't the hardware, which is ordinary; it's that the person chose it, and it answers to them.

That distinction is also the whole investment thesis. In December 2025, SEMCAP Health led a targeted $25 million Series D, the first close of which brought SafeinHome's total funding to $67 million. What's notable is what the investors led with. Not the sensors. Not the AI. Human dignity. In a sector crowded with gadget pitches - fall detectors, panic buttons, wearables that ping a call center - the pitch that scaled was the one about restraint: technology that has your back rather than technology that watches you.

The math that makes Medicaid pay attention

Most "person-centered care" language in healthcare is aspirational and un-priced. SafeinHome's is priced. The company reports a 96% monthly retention rate and a 99% daily support-plan completion rate, alongside that 60% cost reduction per hour. Those three numbers, taken together, are the argument. If people stay, if the plans actually get done, and if it costs meaningfully less than posting a direct-support worker in every home, then the usual tradeoff between quality and cost stops applying. That is precisely the pitch that lands with state Medicaid programs, which fund SafeinHome's service as a Home and Community-Based Service (HCBS) and are, perpetually, short on both money and workers.

SafeinHome sits at an interesting intersection of the direct-care labor shortage. There are not enough support workers, wages are low, turnover is brutal, and demand keeps rising as more people with I/DD and older adults choose to live at home rather than in facilities. One remote staffer, supported by sensors and software, can cover moments across many homes. Whether that fully substitutes for a human in the room is a real question - and SafeinHome is careful to say it doesn't try to. The service is designed to work alongside in-person supports, filling the overnight hours and the in-between moments, not replacing the caregiver entirely.

Fourteen years of data, now aimed at prevention

The quiet asset here is time. SafeinHome has accumulated more than 30 million hours of real-world support data and over 5 million sensor hours. A cynic would call that a moat; the company frames it as raw material for a specific job - spotting the shape of a crisis before it fully forms. The newer platform layers AI over that history to personalize support plans and manage potential crises proactively, which in this context means noticing that a routine has drifted, a night is going wrong, a pattern looks like the early hours of a past emergency. It's a use of AI that is easy to describe and hard to do well, and the value of the dataset is that it's collected in exactly the setting where the predictions have to work.

There is also a plainly sensible design philosophy running through all of it. The sensors are small on purpose. The tablet stays dark until needed on purpose. Support plans are built to shrink as a person gains confidence and skill, rather than expand - the goal being to make the company progressively less necessary to any given individual, which is not a growth strategy anyone teaches in business school but is exactly right for this domain.

Where it goes from here

SafeinHome now operates across 17 states, with California named as its newest market on the back of the Series D. The expansion logic is straightforward: each state Medicaid program is a distinct contract with distinct regulations, so growth looks less like flipping a switch and more like methodically clearing one statehouse at a time. It's slow, unsexy work. It is also, given the demographics and the budget math, work with a very long runway. The company that started in 2012 building home sensors for a market the tech industry mostly ignored has ended up holding a fairly enviable position - useful to families, cheaper for payers, and pointed at a need that is only getting larger.