Kaigo Health Is Betting That Phone Calls Will Save Medicare, and the Math Checks Out

The Macro: Medicare Is Paying for Phone Calls Nobody Is Making

There is a strange thing happening in American healthcare right now. Medicare has built a set of programs that literally pay providers to call patients at home. Remote care management, chronic care management, post-discharge follow-ups. The reimbursement codes exist. The revenue is real. And most providers are leaving that money on the table because they simply do not have the staff to make the calls.

The numbers are hard to ignore. There are roughly 60 million Medicare beneficiaries in the United States. A significant percentage of them qualify for remote care management services. Each enrolled patient generates recurring monthly revenue for the provider. But actually running these programs requires dedicated care coordinators who call patients, ask structured questions, document responses, flag anomalies, and loop in clinicians when something looks wrong. The average care coordinator can manage maybe 200 patients. A mid-size health system with 10,000 eligible patients would need 50 coordinators just to run the program. That is an entire department most systems cannot afford to build.

Meanwhile, readmission rates remain stubbornly high. Patients get discharged, nobody checks on them for two weeks, and they show up in the ER again. CMS penalizes hospitals for excess readmissions. The whole thing is a doom loop: providers need staff to prevent readmissions, they cannot hire enough staff, patients bounce back, and the system pays more.

This is not a technology gap. It is a workforce gap. And that is exactly where AI voice agents start to make a lot of sense.

The Micro: Three Founders Who Noticed the Obvious

Kaigo Health was founded by Peter Chien (CEO), Lukas Klaiber (COO), and Kazuki Shin (CTO). They came through Y Combinator’s Fall 2025 batch with a simple thesis: if the bottleneck in outpatient care is human phone calls, and AI can now make phone calls that sound like actual humans, then you can scale these programs 100x without scaling headcount.

The product is an AI agent that calls Medicare patients on behalf of their providers. It handles post-discharge follow-ups, ongoing chronic care check-ins, and wellness monitoring. The calls are structured around clinical protocols. The AI asks the right questions, listens to the answers, and surfaces actionable data to the care team. If a patient reports worsening symptoms or a medication issue, the system flags it immediately rather than waiting for the patient to remember to call their doctor next week.

The interface choice is deliberate. They are using phone calls, not apps. Not patient portals. Not text messages. Phone calls. Because the target population is Medicare patients, whose average age is north of 70, and who overwhelmingly prefer voice communication over any digital alternative. Every digital health startup that has tried to get elderly patients to download an app and log symptoms daily has learned this lesson the hard way. The phone call is the product.

I find the approach refreshingly honest. Most healthcare AI companies try to wedge themselves into clinical workflows with complex integrations and change management nightmares. Kaigo is inserting itself into a workflow that is already supposed to happen but is not happening because of labor constraints. The providers already want to make these calls. They already get paid when the calls happen. They just need someone (or something) to actually do it.

The team is small, just three people, operating out of the Bay Area. The competitive landscape includes companies like Abridge, which focuses on clinical documentation, and CareHarmony, which does traditional care management outsourcing with human nurses. But nobody has quite nailed the AI-first phone call approach for Medicare remote care management specifically. Signify Health does home visits, not calls. Eko focuses on diagnostic devices. The phone-call-as-interface angle is surprisingly underexplored.

The Verdict

Kaigo Health is going after one of those rare opportunities where the payer (Medicare) is already willing to write checks, the provider is already willing to participate, and the only thing missing is the capacity to execute. That alignment almost never happens in healthcare.

The risk is regulatory. CMS rules around what counts as a billable care management interaction are specific and evolving. If AI-conducted calls do not satisfy documentation requirements or if CMS changes the rules around who (or what) can perform these services, the entire model could get undercut overnight. Healthcare reimbursement is a moving target and startups that depend on specific billing codes need to stay nimble.

I would also watch for clinical validation. The calls need to actually catch problems early, not just generate data. If Kaigo can show measurable reductions in readmission rates and emergency department visits, providers will line up. If the calls feel like robocalls and patients start hanging up, it falls apart fast.

At thirty days, I want to see provider contracts signed and calls being made. At sixty days, I want patient engagement data: how long are calls lasting, how many patients complete them, what percentage generate clinical alerts. At ninety days, the question is whether this is driving real revenue for providers or just creating noise. The thesis is strong. The execution surface is narrow and well-defined. For a three-person team, that is exactly where you want to be.