Hippocratic AI's safety-first agents bet: $9-an-hour AI nurses, $3.5B valuation

Builds safety-focused generative AI voice agents for healthcare's routine work: chronic-care check-ins, medication onboarding, colonoscopy prep and pre-operative calls, sold to hospitals and payers.

Health systems would pay for AI agents handling routine, non-diagnostic patient contact — and refusing to diagnose would be the feature that gets past regulators.

$404M total: $141M Series B at $1.64B in early 2025, then a $126M Series C led by Avenir Growth at $3.5B with CapitalG, General Catalyst, a16z, Kleiner Perkins and Universal Health Services (November 2025)

Hippocratic AI was co-founded in 2023 by CEO Munjal Shah with physicians, hospital administrators and AI researchers drawn from El Camino Health, Johns Hopkins, Stanford, Microsoft, Google and NVIDIA. The founding insight: most healthcare AI aimed at diagnosis, the highest-risk task, while hospitals were drowning in low-risk, high-volume patient contact that nobody covered.

The company leaned into the paradox of 'safety-first AI nurses': in March 2024 it announced an NVIDIA-powered roster of voice agents — patients pick a persona, 'engaging' or 'direct' — delivering colonoscopy prep and loneliness companionship at $9 an hour, a price point that made national news and angered parts of the nursing profession even as 40+ healthcare companies began testing. The design constraint stayed firm: agents handle non-diagnostic work and hand off anything clinical. Health systems converted: Cleveland Clinic, Northwestern Medicine, Ochsner Health, Moffitt Cancer Center and University Hospitals among 50+ partners by late 2025. In November 2025 Avenir Growth led a $126M Series C at $3.5B — near-double the January 2025 valuation — with CapitalG, General Catalyst, a16z, Kleiner Perkins and hospital operator Universal Health Services joining, bringing total funding to $404M.

US healthcare's binding constraint is staff time, not software budgets, so automating low-risk patient contact has a direct, quantifiable labor payback for health systems.

The non-diagnostic scope was a regulatory and trust moat: it kept Hippocratic out of FDA device territory and gave risk-averse hospitals a defensible yes.

Pricing agents like labor ($9 an hour) rather than like software made the value proposition legible to hospital CFOs comparing against $40-plus hourly nursing costs.

Clinician co-founders and named safety research turned the most skeptical buyer — hospital clinical leadership — into the channel.

In high-stakes industries, refusing the glamorous task is the wedge: Hippocratic won health systems by automating the low-risk volume everyone had and avoiding the diagnosis nobody would let AI touch.

As of November 4, 2025, Hippocratic AI counts 50+ health-system partners and a $3.5B valuation on $404M raised, and says the Series C funds mergers, product development and international growth. The company has kept its agents out of diagnosis; the open question its critics raise — whether low-risk voice AI stays low-risk at scale — moves with every new deployment.