Caregiver shortages, margin compression, regulatory complexity, and a wave of private equity consolidation are converging at once. No doubt, post-acute and home care is facing a reckoning. The providers that will thrive in the decade aren't the ones adding headcount. They're the ones rethinking the operating model entirely.
That's the vision driving Kunal Sarda, CEO and co-founder of Arya Health, and it's why we're proud to welcome him as a keynote speaker at this year's Home Care Innovation Forum. Arya Health builds AI-powered digital agents for the at-home care industry so that the humans working in it can focus on what matters most: caring for clients.
The results are hard to ignore. Locations using Arya's platform see an average of 25% more hours filled. Onboarding time has been cut in half. And the company has grown revenue more than 6x, backed by leading investors including executives from OpenAI.
Kunal is a compelling voice in the industry, not just because of his company’s growth story, but the clarity around what's broken in at-home care administration, and what it’ll take to fix it.
We sat down with Kunal to get his take on AI's role in the future of care in the home, where the line is between automation and human connection, and what the best operators are doing right now to pull ahead.
In 5 years, post-acute care administration shouldn’t feel like “administration” at all.
Today, that 25 cents on the dollar is a tax on the system. It’s manual coordination, fragmented workflows, constant back-and-forth, and teams spending their time keeping the system running
instead of delivering care.
If AI agents reach their full potential, that layer gets fundamentally rebuilt. Scheduling happens in real time, continuously optimizing across patients, caregivers, and constraints without human intervention. Compliance is no longer reactive and audit-driven. It’s embedded, always-on, and handled in the background. Documentation, intake, billing, recruiting - all orchestrated end-to-end by agents that understand context, take action, and coordinate across systems.
Instead of dozens of disconnected tools and large back-office teams stitching things together, you have an autonomous operational layer that runs the business alongside clinical teams.
The impact is not just cost reduction, though that’s meaningful. It’s a structural shift in how these organizations operate.
Providers can scale without adding administrative headcount. Caregivers spend more time on patients and less time on logistics. Response times shrink from hours to seconds. Entire workflows that used to require coordination across multiple people just... happen.
And most importantly, the economics of delivering care outside the hospital start to work again. That’s the real vision. Not just making admin more efficient, but removing it as a constraint on access, quality, and growth.
The caregiver shortage isn’t just a supply problem. It’s a system problem. Two metrics tell the whole story:
* ~50% of caregivers leave within the first 90 days.
* often takes weeks for a newly hired caregiver to get staffed on their first patient
That combination creates a vicious cycle. Agencies are constantly recruiting to replace churn, while also losing people before they even start. Recruiting teams end up running in place, trying
to fill a leaky bucket.
At the same time, tools like Indeed Quick Apply have made it easier than ever to apply. That’s increased volume, but not quality. Recruiters are buried in applicants, trying to find the right caregiver in a sea of noise. Traditional workforce strategies don’t fix this because they don’t change the underlying math. AI agents do.
First, they automate the top of the funnel. Instead of humans manually reviewing hundreds of applicants, agents can vet, qualify, and prioritize candidates in real time, so recruiting teams focus only on the right people.
Second, they compress the time from hire to first shift. This is critical. Agents coordinate onboarding, credentialing, and scheduling immediately, so caregivers get to work faster and don’t drop off before they ever start.
Third, they ensure better long-term matches. By continuously understanding each caregiver’s preferences, availability, and work patterns, agents can staff them into the right cases, not just any open shift. That leads to more stable schedules and higher retention.
This is where recruiting and scheduling stop being separate functions and start working as one system. That’s exactly what we’re building at Arya.
Our recruiting and scheduling agents work hand in hand to identify the right candidates, move them through the funnel, get them staffed quickly, and keep them matched in a way that works for both the caregiver and the agency. The result is a fundamentally more stable workforce in the long term.
Automation in home-based care has to be approached with a clear principle: It should strengthen relationships, not replace them. This industry runs on trust. Between office staff and leadership. Between caregivers and coordinators. Between patients and providers. If automation erodes that trust, it fails no matter how efficient it is.
There are three relationships that matter most:
This is often the most overlooked group. They’re the ones closest to the operational pain, but also the most anxious about what AI means for their jobs. There’s real fear, and real inertia.
Leadership has to address this head-on. The question every staff member is asking is simple: what’s in it for me?
If automation is positioned as cost-cutting or headcount reduction, it will be resisted. If it’s positioned as a way to remove the most repetitive, frustrating parts of the job and allow people to focus on higher-value work, it gets a very different response.
The goal is not to replace office staff. It’s to elevate them. Less time chasing down paperwork, making repetitive calls, or fixing schedule gaps. More time solving problems, supporting caregivers, and improving patient experience. That shift only works if leadership is explicit about job security and intentional about how roles evolve.
The patient-provider relationship in home care is already fragile. It’s built on consistency, responsiveness, and trust. Automation cannot get in the way of that. Done right, it actually strengthens it.
Faster responses. Fewer missed visits. Better coordination. More proactive communication. Patients feel seen and supported, not because they’re talking to more people, but because the system around them works better. The key is that automation handles the operational burden behind the scenes, while the human moments stay human.
Patients shouldn’t feel like they’re interacting with a system. They should feel like their provider is simply more reliable, more attentive, and more in sync with their needs.
Caregivers are often stuck in transactional interactions. Back-and-forth about schedules. Last-minute changes. Constant coordination with the office. It’s frustrating, and it pulls them away from what they actually care about, which is caring for patients. Automation should remove that friction.
If done well, scheduling becomes smoother, communication becomes clearer, and the day-to-day logistics fade into the background. That creates space for more meaningful relationships between caregivers and coordinators. Conversations shift from “Can you cover this shift?” to “How is this case going?” or “What do you need to succeed?”
In other words, automation reduces transactional interactions so that relational ones can increase. That’s the bar. Not just doing things faster or cheaper, but redesigning the system so that every stakeholder spends less time on the mechanical and more time on what actually matters.
Ensuring end-users stay at the center of AI in healthcare comes down to two things:
1. Experience: Our belief is simple: AI should fit the existing flow of care, not force people into a new one. Patients and caregivers already operate in a rhythm, how they communicate, how they receive updates, and how they interact with the office. Introducing new apps, portals, or workflows often creates more friction than value.
That’s why we think of Arya as a single pane of glass, one that sits naturally inside the channels people already use, primarily voice and text. No logins to remember. No new systems to learn.
If AI adds friction to the day-to-day experience of a caregiver or patient, it’s a failure. The bar is that things should feel simpler, faster, and more intuitive without people having to change how they work.
2. Outcomes: Experience matters, but outcomes matter more. For patients, the goal is continuity of care. For caregivers, it’s continuity of employment. That means AI cannot be optimized for tasks in isolation. It’s not just about filling a shift. It’s not just about updating an expired license. It’s about making sure the right caregiver is matched to the right patient from the start, based on preferences, needs, and context and then maintaining that match over time. It’s about understanding caregivers more deeply so agencies can meet their needs, create stable schedules, and reduce churn.
When AI is designed this way, it improves the lived experience of care for both sides. That’s how you keep end-users at the center, not by building more technology, but by making sure the technology disappears into better experiences and better outcomes.
What much of the industry hasn’t fully internalized yet is that AI is going to redefine homecare operations at a much more structural level than simply automating tasks.
Today, most organizations are approaching AI through the same lens they used for SaaS. They are looking for ways to improve narrow slices of work, better scheduling tools, better recruiting pipelines, more efficient compliance tracking. This approach assumes that the underlying structure of work remains intact, and that technology’s role is to make each function incrementally better. That assumption breaks with AI.
The next generation of home care operations will not be organized around functions. They will be organized around journeys, both for caregivers and for patients. The work that exists today in silos will begin to collapse into unified, continuous processes.
On the caregiver side, recruiting, scheduling, and compliance are currently treated as distinct domains with separate teams and separate incentives. Over time, those boundaries will erode. AI agents will handle much of the execution across these domains, and human operators will shift toward owning the full caregiver lifecycle. The focus moves from completing tasks within a function to managing outcomes across recruiting, onboarding, staffing, and retention as a single system.
A similar shift happens on the patient side. Intake, scheduling, coordination, and billing are fragmented steps today, often requiring multiple handoffs and systems. In the future, these will converge into a single, end-to-end process of client management. AI will orchestrate the flow across these stages, allowing organizations to deliver a more continuous and responsive experience.
The implications for the org chart are significant. This is not just about efficiency gains or cost reduction. It changes how teams are structured, how roles are defined, and how performance is measured. Fewer people will be responsible for broader scopes of work, with AI handling the operational complexity that previously required entire departments.
It also changes how providers need to think about technology.
Solutions that focus on optimizing individual functions will struggle to support this shift because they reinforce the very silos that are beginning to disappear. The systems that will matter are those that can operate across the full caregiver and patient journey, enabling organizations to rethink how work is done rather than simply improving how it is done today. This is the transition underway. Not incremental improvement, but a fundamentally new operating model for home care.