Neil Batlivala spent seven years building Pair Team, a healthcare company unknown to most of Silicon Valley. It focused on patients with chronic conditions often overlooked. Unstable housing. Food shortages. No reliable transportation. About a third of Americans face such barriers.
Last month everything changed. Pair Team announced its acceptance into a new Medicare program. One of just 150 participants. The initiative launches July 5. It tests what happens when federal payments reward results instead of tasks.
TechCrunch first highlighted how this model aligns incentives for artificial intelligence in ways traditional Medicare never could. The program, called ACCESS, stands for Advancing Chronic Care with Effective, Scalable Solutions. It runs 10 years. Its payment approach ties full reimbursement to measurable patient improvements. Lower blood pressure. Better blood sugar control. Reduced pain or depression scores.
Conditions covered include diabetes, hypertension, chronic kidney disease, obesity, depression and anxiety. These affect more than two-thirds of Medicare beneficiaries. The official CMS description emphasizes technology-supported care. AI diagnostics that flag at-risk patients. Devices tracking biomarkers. Software streamlining workflows. All guided by clinicians yet flexible enough for virtual, asynchronous or in-person delivery.
Traditional Medicare pays for time with a clinician. Minutes logged. Visits completed. No direct mechanism exists for an AI agent that checks in daily. Coordinates housing help. Reminds patients about medications. Or simply listens. ACCESS changes that. Organizations receive predictable recurring payments for managing a patient’s condition. They keep the full amount only by hitting outcome targets across their population.
“The government is creating swim lanes for AI innovation in traditionally regulated industries,” Batlivala told TechCrunch. “The best solution wins, which, in regulated industries like healthcare — that’s not been the case.”
But most of the broader tech world remains unaware. Digital health funding reached its highest first-quarter total since the pandemic this year. AI companies took the lion’s share. Yet this Medicare shift has registered mainly in specialized health tech circles.
Pair Team started in 2019 with a simple premise. Health outcomes improve only when care addresses the full context of a person’s life. The company now employs roughly 850 clinical professionals. It operates what it calls California’s largest community health workforce. Revenue exceeds nine figures. It has raised about $30 million from investors including Kleiner Perkins, Kraft Ventures and Next Ventures.
Evidence backs the approach. A peer-reviewed study in the Journal of General Internal Medicine, co-authored by Pair Team researchers, showed strong patient engagement. It also delivered sharp drops in avoidable emergency and inpatient use. One in four hospital visits. One in two ER visits. Those numbers disappear under the company’s care, Batlivala claims.
For years that model relied heavily on humans. Scaling proved expensive and slow. Nine months ago Pair Team introduced Flora. A voice AI agent. Available 24 hours a day. It handles intake. Coordinates referrals. Performs check-ins. One 67-year-old woman living out of her car, managing PTSD and congestive heart failure, spoke with Flora for over an hour. “It was both incredible and depressing,” Batlivala recalled. “Flora was probably the only ‘person’ she’d talked to in weeks about her situation.”
Such extended conversations have become routine. Companionship itself functions as an intervention. The economics shift dramatically with AI handling much of the patient interaction. CMS pays less per patient per month than many expected. That lower rate, Batlivala argues, forces efficiency. “If you want to build a model that truly incentivizes the use of AI, the reimbursement rates have to be low. The economics only work if you’re running a lean, AI-first operation.”
The architects behind ACCESS bring startup experience. Abe Sutton directs the CMS Innovation Center. He previously worked as a venture capitalist at Rubicon Founders. Jacob Shiff serves as chief AI and technology officer. A former healthcare founder himself. Their backgrounds show. The model features outcome-based payments. Direct-to-consumer enrollment. A clear push for competition among diverse participants. AI doctor startups sit alongside virtual nutrition providers, device makers and even wearable companies like Whoop.
Batlivala expresses skepticism about some entrants. “I’m a big fan of wearables, but for a senior who’s struggling with food insecurity, I don’t know how much Whoop is going to be able to do.” Pair Team, by contrast, built toward this moment for more than five years.
CMS describes the model in detail on its site. Participants must be Medicare Part B-enrolled organizations. They designate a physician as clinical director. Patients enroll voluntarily, often after referral from their regular provider. Outcomes are risk-adjusted and published transparently. Beneficiaries and referring clinicians gain data to choose high performers. The agency stresses accountability. HIPAA compliance. FDA requirements for devices and software. Secure, interoperable systems using CMS APIs.
Four clinical tracks organize the conditions. Early cardio-kidney-metabolic issues such as hypertension and prediabetes. More advanced cardio-kidney-metabolic disease including diabetes and chronic kidney disease. Musculoskeletal pain. Behavioral health conditions like depression and anxiety. Each track carries specific measures and targets. Improvement in biomarkers. Validated patient-reported outcomes for pain, mood and function.
Payments begin with an initial care year. An optional continuation period follows at a reduced rate. Rural patients receive a fixed adjustment. Full payment depends on the share of patients meeting targets. The threshold rises each year. Prevention sits at the center. Organizations succeed by keeping people healthy rather than treating sickness after it escalates.
Risks remain real. Participants feed sensitive data — conversations about housing instability, mental health, chronic disease — into federal systems. CMS has a documented history of breaches, including exposed Social Security numbers. A 2023 Congressional Budget Office analysis found the Innovation Center increased federal spending by $5.4 billion in its first decade instead of delivering expected savings. Past performance offers no guarantee.
Pair Team holds partnerships reaching roughly 500,000 potential patients. It aims for one million within three years. The company believes the lower reimbursement rates actually favor AI-heavy models. Humans cost more. Automation scales.
This Medicare experiment arrives amid broader shifts. The WISeR model, launched earlier in 2026, uses AI to screen prior authorization requests for certain services in six states. It targets procedures with little proven benefit for some patients. Critics worry about delays and bias. Yet it signals the same direction. Technology will play a larger role in payment decisions.
Healthcare investors watch closely. So should the rest of the tech sector. Payment models dictate adoption speed. When reimbursement aligns with outcomes that AI can influence — early detection, continuous monitoring, personalized follow-up — incentives realign. Companies must prove not just technical accuracy but measurable impact on cost and health status.
ACCESS won’t transform Medicare overnight. It’s voluntary. Limited to Original Medicare for now. The first cohort tests the waters. Success could lead to wider adoption. Medicare Advantage plans. Medicaid. Even commercial insurers. The CMS blog post calls it “a major step forward in aligning payers, providers, and patients around what matters most: better health at an affordable cost.”
Batlivala sees it as validation. Years of work in the shadows suddenly sit at the center of federal policy. “It’s a payment model transformation,” he said. “You just couldn’t do this before.”
Whether the broader tech industry wakes up to this opportunity remains an open question. The infrastructure now exists. The incentives point toward AI that delivers results rather than novelty. The patients most in need wait in the background. For now the experiment begins. Outcomes will decide the winners.
WebProNews is an iEntry Publication