The sound of a busy medical office is familiar to almost anyone who has tried to get an appointment: the endless ringing, the automated voice listing options that never quite match your problem, the hold music that pipes in after you press the wrong button by accident. For many patients, the simple act of seeing a doctor has become a test of patience and persistence. For health systems, the frustration is just as real, because behind every unanswered call and abandoned web form lies an empty exam room and a lost opportunity to care. This is the messy, unglamorous world that Seattle-based DexCare has been quietly working on for years. On September 21, 2026, the company made a consequential move, announcing that it would acquire Mila Health, another Seattle-area startup that has built AI agents capable of calling and texting patients to help them secure medical appointments. The purchase price was not disclosed, but the strategic logic was clear. DexCare brings the plumbing and infrastructure of healthcare scheduling, while Mila brings a conversational, artificial intelligence-powered front door. Together, they want to do more than just make the phone ring a little less. They want to reimagine the entire experience of getting care, using AI that is reliable, safe, and human enough to make patients feel heard rather than processed.
To understand why this deal matters, it helps to understand what DexCare actually does. The company did not start out as a glamorous startup with a giant screen and a catchy demo. Instead, it came from inside one of the largest health systems in the country. Providence, the massive nonprofit healthcare organization, launched DexCare out of its digital innovation group about a decade ago. The initial problem was practical: how do you get patients to the right doctor at the right time, especially as more people began looking for care online? The answer, it turned out, was not just a website with a scheduling button. Health systems have to manage an enormous amount of information before someone can be booked. Does the patient have the right insurance? Does the referral check out? Is the specialist accepting new patients? What are the specific rules for a given clinic? DexCare became a kind of operating system for healthcare access, helping organizations coordinate their capacity and fill appointment slots in a way that is efficient and clinically sensible. Today, the company says it serves 57 million patients nationwide. Its customer list includes some of the most recognizable names in American healthcare: Kaiser Permanente, Piedmont, Texas Health Resources, and Tampa General. That kind of adoption does not happen overnight. It takes years of building trust and integrating into the messy realities of hospitals and clinics. DexCare has raised $146 million in total funding, including a $75 million round in 2023, and was a finalist for Health Innovation of the Year at the 2023 GeekWire Awards. It also ranks No. 43 on the GeekWire 200, an index of the Pacific Northwest’s top startups, and employs about 133 people, according to PitchBook. In other words, this is a company with serious scale, serious backing, and a track record of actually doing the difficult, unglamorous work of fixing the healthcare system piece by piece.
Mila Health, by contrast, is a much younger and leaner outfit. It was founded in January 2024 by Shailu Verma, a former product leader at UnitedHealth Group and a longtime Amazon Web Services product manager. The company’s bet was that AI could do something very specific and very useful: hold real conversations with patients over the phone or through text messages to answer questions and schedule care. This is not a chatbot that forces you to type symptoms into a box and then forwards you to a human. Mila’s technology is designed to be embedded directly into the systems that healthcare providers already use. The startup says it takes only hours to launch its AI agents, not the months or years that typical software integrations require. That speed is important because healthcare organizations are desperate for help. Front desk staff are overworked. Call centers are understaffed. Patients are frustrated. The labor shortage in healthcare is not just about doctors and nurses; it is also about the administrative workers who keep the machine running. Mila’s AI agents can take over the repetitive, high-volume, low-judgment work of answering appointment inquiries, checking availability, and confirming visits. The technology also understands the nuance that makes healthcare scheduling so hard. It can ask whether a patient has a referral, whether they are a new or returning patient, and whether the visit is routine or urgent. It can handle a conversation the way a well-trained human scheduler would, but without needing to sleep, eat, or take breaks. The company has 25 employees and, notably, appears to have not raised any venture capital at all. It participated in the Advanced Technology Development Center accelerator, an Atlanta-based program, but no equity or funding was exchanged. That is a striking reality in a world where barely a month goes by without another AI startup raising tens of millions of dollars. Mila seems to have built something valuable through focus and talent rather than endless fundraising.
The acquisition makes the most sense when you consider what each company says about the other. Matt Blosl, DexCare’s CEO, captured the mood of the industry when he said that AI solutions are everywhere, but health systems quickly realize that building the infrastructure, scaling the pilot, and managing the agent’s boundaries is untenable. In other words, plenty of vendors are willing to sell a clever AI demo, but very few are honest about what it takes to make that demo work in a real hospital. The challenges are not just technical. They are operational, regulatory, and cultural. Can the AI be audited? Who is accountable if it makes a mistake? How does it behave when a patient is anxious, confused, or in pain? How do you ensure that the AI does not do something that a human scheduler would never do, like promising a patient a type of visit that requires prior authorization? Mila’s co-founder and CEO, Shailu Verma, argues that healthcare cannot rely on out-of-the-box large language models because they are unpredictable and cannot be audited. That is a bold statement in an era when many people assume that general-purpose AI can be pointed at any problem and somehow produce a good answer. But healthcare is different. A hallucinating chatbot that writes a wrong poem is a mild embarrassment. A hallucinating healthcare AI that tells a patient to show up for a surgery without an approved referral is a liability. That is why Mila built its agents to be configurable, with boundaries set by the organizations that deploy them. With DexCare, Verma says, providers and payers can configure an AI agent to complete the tasks they cannot staff for. The acquisition, then, is not just about acquiring new technology. It is about pairing that technology with the infrastructure, governance, and customer relationships that make adoption possible in the first place.
There is also a human story behind this deal. Shailu Verma has spent his career at the intersection of healthcare and technology. At UnitedHealth Group, he served as chief product officer of strategy and innovation, a role that required understanding the realities of payers, providers, and patients. Before that, he spent five years at Amazon Web Services, where he learned how to build scalable, reliable technology. That combination is not common. Many health-tech founders come from either healthcare or tech, but rarely both. The fact that Mila was able to attract and retain a team of 25 people without raising venture capital suggests that the founders were not chasing hype. They were trying to solve a problem. DexCare, meanwhile, has its own founding story that is deeply rooted in the real world of healthcare delivery. Providence created the company because it needed to improve digital patient acquisition, and that origin still shapes how it operates. The company also previously acquired Womp, a Bellingham, Washington-based online commerce platform, in 2022, showing that it is not afraid to buy and integrate external technology when it strengthens its core offering. With this acquisition, DexCare gets a team of people who understand AI deeply and a product that can make its existing platform smarter and more conversational. Mila gets something just as valuable: the ability to reach 57 million patients through established health system partners. The deal also signals that the Seattle startup ecosystem remains a vital place for health innovation. It is perhaps no accident that both companies are rooted in the Pacific Northwest, an area known for its engineering talent, its cloud computing expertise, and its willingness to tackle difficult problems in fields like healthcare.
Looking ahead, the real test will not be whether the two companies can combine their technologies, but whether they can deliver a better experience for everyone involved. For health systems, the promise is clear: fewer empty slots, fewer overtime hours for schedulers, lower administrative costs, and a way to respond to patient inquiries at two in the morning without hiring a night shift. For patients, the hope is more modest but no less important: a response that comes quickly, a tone that feels respectful, and a confirmed appointment instead of a voicemail that never gets returned. There are, of course, legitimate concerns about AI in healthcare. People worry about privacy, bias, and the erosion of human touch. They worry that talking to a machine will feel cold or impersonal. But the companies involved seem to understand that AI must operate within limits, and that the technology should handle the routine so that humans can handle the meaningful. The quote from Verma about auditable AI is not just a technical requirement; it is a promise that healthcare decisions will remain transparent and accountable. Blosl’s comment about managing the agent’s boundaries is a recognition that even the smartest machine needs clear rules. In the end, this acquisition is a quiet reminder that the most successful healthcare innovations are often the ones patients never see. They are not apps with glowing reviews or gadgets with flashing sensors. They are the systems that make a phone call a little easier, a waiting room a little less crowded, and a diagnosis a little faster. By joining forces, DexCare and Mila are betting that AI can be more than a buzzword. They are betting that it can be the helpful voice on the other end of the line, the one that finally says, “Yes, we have an opening Thursday morning,” and means it.
