← February 13, 2027 edition

beacon-health

AI employees for primary care that work inside your EHR without changing your workflow

Beacon Health Puts AI Employees Inside Primary Care Clinics and Asks Doctors to Change Nothing

HealthcareArtificial IntelligenceValue-Based CarePrimary Care

The Macro: Primary Care Is Drowning in Admin Work

Two hundred million Americans rely on primary care as their main source of healthcare. The doctors managing those patient panels are stretched impossibly thin. Between preventive screenings, prior authorizations, referrals, risk adjustment coding, and quality measures, the administrative work has swallowed the clinical work whole.

Value-based care was supposed to fix this. Instead of paying per visit, insurers pay for outcomes. In theory, this rewards prevention and comprehensive patient management. In practice, it means primary care practices need to track dozens of quality metrics, close care gaps, code patient risk accurately, and document everything. Most practices do not have the staff for this. The result is money left on the table and patients falling through the cracks.

The healthcare AI market is crowded with companies promising to automate clinical workflows. Abridge handles visit documentation. Nabla does ambient listening. But most of these tools require physicians to change how they work. Beacon Health, backed by Y Combinator, takes a different angle: their AI workers operate inside existing EHR systems like invisible employees, handling the back-office grind without asking anyone to change their process.

The Micro: Just Plug It In

Mark Pothen (CEO) grew up in his mother’s primary care practice and spent six months embedded in one managing operations before starting Beacon. Obinna Akahara (CTO) built production AI systems across healthcare and enterprise software, with a physics degree from UT Austin. The combination of clinical operations knowledge and engineering chops is exactly what this kind of product needs.

The pitch is dead simple: “Do nothing, change nothing, make money.” Beacon’s AI workers connect to a practice’s EHR and start identifying revenue opportunities in value-based care contracts. They find unclosed care gaps, optimize risk adjustment coding, and handle Annual Wellness Visit outreach. The practice does not install new software, learn new interfaces, or change workflows.

This is a smart positioning choice. Healthcare IT adoption is slow because clinicians are overworked and resistant to new tools that add friction. By making the product invisible to the end user, Beacon sidesteps the biggest barrier to adoption in health tech.

The success-based pricing model is also worth noting. Rather than charging a flat SaaS fee, Beacon ties its revenue to the additional revenue it captures for the practice. This aligns incentives cleanly. If the AI does not find money, the practice does not pay.

The competitive set includes companies like Innovaccer, Aledade, and Pearl Health, all working on value-based care enablement. But most of those platforms require significant implementation effort and workflow changes. Beacon’s “zero-lift” approach is a meaningful differentiator if it delivers.

The company already has customers across multiple practices including MWA, Kaaya Health, and others. The HIPAA-compliant infrastructure is table stakes but necessary.

The Verdict

Beacon Health is making a bet that primary care practices will pay for revenue they did not know they were missing, as long as finding it requires zero effort on their part. I think that bet is correct.

At 30 days: what is the average revenue uplift per practice, and how quickly does the AI identify its first actionable care gap after connecting to the EHR?

At 60 days: are practices renewing after the initial engagement, or is the low-hanging fruit getting picked clean in the first month?

At 90 days: can Beacon expand beyond risk adjustment and AWV outreach into other administrative workflows like referral management and prior authorizations?

The “change nothing” positioning is Beacon’s superpower. If the AI consistently finds revenue that practices were leaving behind, this becomes an easy yes for every primary care group in the country. The question is whether the revenue capture is large enough and consistent enough to build a big business on top of.