Ortet Launches With $500 Million From Thoreau to Build a Frontier AI Lab for Health

Ex-Genentech, Meta and AWS AI veterans have launched Ortet, a New York health AI lab, with a $500 million commitment from Matt Holt's Thoreau.

A new artificial intelligence company called Ortet has launched with a $500 million commitment from the investment firm Thoreau, billing itself as a "frontier AI lab for health" — one of the largest single backings for a healthcare-focused AI startup to date.

The launch was announced Tuesday in a press release covered by Endpoints News, Fierce Healthcare, Axios Pro and the San Francisco Business Times. According to Axios, Thoreau is the investment firm led by Matt Holt, the former president of private equity at New Mountain Capital.

If you follow the latest AI developments, healthcare has quietly become one of the most contested frontiers for frontier-scale AI investment — and Ortet's debut is the newest evidence of that shift.

Who Is Behind Ortet

Ortet is led by Kyunghyun Cho, an NYU professor and well-known deep learning researcher whom the company's press release describes as an AI pioneer. Endpoints News reports that Cho is among a trio of former AI scientists and engineers at Genentech, Roche's biotech arm, who co-founded the company, with veterans of Meta and Amazon Web Services also joining the effort.

The combination is deliberate: the founding team pairs machine-learning researchers who have worked at the frontier of the field with people who have spent years applying AI inside drug development and cloud-scale infrastructure. The San Francisco Business Times reports the company is based in New York.

Cho is not a newcomer to applied AI research — he has been a prominent academic voice in deep learning for over a decade, and Endpoints notes he spoke at its own AI Day conference in 2024 about the intersection of machine learning and medicine.

What Ortet Plans to Build

According to the launch announcement, Ortet will build a patient-centered foundation model for health, supported by integrated research, data and compute capabilities. The company describes its ambition as a full-stack platform for health — foundational compute, rules and research infrastructure — rather than a single product.

Thoreau's announcement, as summarized by Healthcare Deal Hub, says Ortet develops AI models and infrastructure to address healthcare challenges across biological, clinical, operational and financial dimensions. That framing matters: most healthcare AI startups attack one slice of the system, such as clinical documentation or drug discovery, while Ortet is positioning itself as vertical infrastructure spanning the entire industry.

The "frontier lab" model comes to healthcare

The more significant signal in Ortet's launch is structural. Frontier AI labs — the organizations training the largest models — have historically been general-purpose companies like OpenAI, Anthropic and Google DeepMind. Ortet is adopting that playbook, including the capital intensity, but pointing it at a single vertical from day one.

Whether a vertically focused lab can compete for talent and compute with the general-purpose giants is an open question. But healthcare is arguably the domain where a purpose-built foundation model has the clearest moat: clinical data is heavily restricted, regulatory requirements shape what models can touch, and the difference between a general chatbot and a clinically grounded system is the difference between a demo and a product.

Why Investors Are Betting Big on Health AI

Thoreau's half-billion-dollar commitment is a wager that healthcare's structural problems — labor shortages, administrative waste and the slow pace of translating research into care — are exactly the kind of problems large models can address, provided someone builds with the domain's constraints in mind rather than around them.

Holt's background running private equity at New Mountain Capital, a firm with a long history of healthcare investing, suggests the commitment is a strategic platform bet rather than a typical seed check. Axios reported the commitment is being used to build out the company's compute, research and infrastructure foundation from the start.

For context on the broader wave: health systems and insurers have been rapidly adopting AI tools, regulators have begun drafting rules specifically for clinical AI, and every major frontier lab now lists healthcare among its priority verticals. Ortet's founders are betting that a company designed for healthcare from the ground up will out-execute general-purpose labs dabbling in the domain.

What to Watch

Ortet launched without disclosing clinical partnerships, regulatory strategy or a product timeline, so the immediate roadmap is thin on specifics. The milestones that will matter: first published research from the foundation model effort, healthcare system pilot deployments, and any engagement with the FDA or equivalent regulators — all standard markers for whether a health AI lab is converting capital into clinical reality.

With $500 million committed at launch, Ortet has bought itself the compute and the calendar to attempt something few startups try: building frontier-scale AI for a single, heavily regulated industry. Whether that focus is an advantage or a constraint will become clear over the next few years.