SAN FRANCISCO—In an overcrowded, standing-room-only room on the 32nd floor of the Westin St. Francis, Daniel Nadler, Ph.D., one of the co-founders of OpenEvidence, made the case for developing “medical super-intelligence” based on agentic artificial intelligence that is multimodal and multicloud.
It's a future that OpenEvidence, a fast-growing health tech company that developed an AI-powered medical search engine, is already working to build, he asserted.
Advances in agentic AI will enable a system of AI agents that can be subspecialists in clinical areas, he told investors and attendees during a packed session at the 2026 J.P. Morgan Healthcare Conference.
OpenEvidence developed an AI-powered medical search engine and generative AI chatbot exclusively for doctors that summarizes and simplifies evidence-based medical information. The company has seen rapid growth and is a widely popular tools among clinicians. It is now actively used daily, on average, by more than 40% of physicians in the U.S., spanning more than 10,000 hospitals and medical centers nationwide, according to the company.
In December alone, OpenEvidence supported about 18 million clinical consultations in the U.S., Nadler said.
The startup was launched by Nadler, who previously founded Kensho Technologies, which was sold for $700 million to S&P Global, and Zachary Ziegler, OpenEvidence's chief technology officer, to make it easier for clinical specialists to access medical and scientific evidence to help with clinical diagnoses.
The company has strategically differentiated itself from other medical AI chatbots by being an early mover and collecting tens of millions of clinical consultations as well as through its content partnerships with medical journals that represent the gold standards of medical knowledge, Nadler noted in an interview back in July.
OpenEvidence has formed strategic content partnerships with the American Medical Association, The New England Journal of Medicine, the Journal of the American Medical Association and all 11 JAMA specialty journals including JAMA Oncology and JAMA Neurology.
Nadler credits the company's approach to partner with medical journals and medical organizations, or what he referred to as being similar to Apple iTunes' approach to music, as key to the company's market traction and ongoing success.
The three-year-old company's valuation hit $6 billion in October after it banked $200 million in series C funding, just three months after it raised $210 million in a series B. The startup has raised nearly $500 million since its founding in 2022. Major backers include Sequoia, Google Ventures, Kleiner Perkins, Coatue, Thrive, Conviction, Iconiq, Greycroft, Breyer Capital, Bond, Blackstone and Mayo Clinic.
OpenEvidence now has set its sights on building specialist AI models to evolve its platform.
“Super-intelligence would be something like, in any given field, whether it's law or engineering or medicine, you have a system that can achieve expert-level knowledge and understanding in every sub-domain of the field. But how would you build it? And the answer is, you would build it in the same way as a hospital is built in the case of medicine, or in the same way that in the case of engineering, NASA or SpaceX is organized,” Nadler said.
At leading hospitals and academic medical centers, a team of specialists and subspecialists works together to care for patients, such as a team that is comprised of oncologists, pathologist and radiologists, he noted.
For a patient with multiple conditions, the optimal approach would be specialists, such as neurologists, dermatologists and general practitioners or family physicians, who “understand the interplay” of the patients’ conditions and treatments.
“You can't just get that from any one lookup. There's no sort of search that's going to solve that. What you would need is to train a neurologist AI and train a dermatologist AI on neurological reasoning or dermatological reasoning to encapsulate and distill the thought process of how a specialist or subspecialist in the field would go about reasoning through a given question,” he said.
The future of agentic AI for healthcare could entail a digital twin neurologist interacting with the digital twin of the dermatologist, he said. “You would want these two AIs to actually have a discussion, have a deliberation” about the patient’s treatment, he said.
“If this future gets built out, you can imagine a world where every patient and their GP, or whoever is on the intake side, is the front door to an ensemble of digital twins, basically AI sub-experts trained, each one of them, in a very, very different way,” he noted.
Nadler asserted that the first steps toward this future could be in oncology AI, pointing to OpenEvidence’s partnership with the National Comprehensive Cancer Network (NCCN). The two organizations announced a licensing agreement in November to make the NCCN’s oncology guidelines accessible through OpenEvidence’s medical AI tech.
“We’re training an AI that is not a generalist AI in the way that any LLM is, RAG or non-RAG (retrieval-augmented generation). We're training, with NCCN, an AI that is optimized around the objective function of oncological reasoning in very specific clinical contexts. The hard part of the technical work once we do that is the sort of architecture that I've outlined here, which is, how do you get these subspecialist AIs to collaborate on a real patient case?” Nadler said.
An “ensemble" agentic AI architecture could be significantly beneficial to smaller medical practices and providers in rural areas that are often left out of large, enterprise AI integrations that target the bigger health systems, according to Nadler.
“Roughly half of doctors the United States practice in medical practices of 10 or fewer physicians; many of them practice in rural areas, underserved areas,” he said, calling out an email the company received from the medical director of a cancer center in rural Georgia calling OpenEvidence "an incredible lifeline for daily practitioners."
“If you do this sort of system that I'm talking about, and you roll this out across the country, and you roll this out in a very organic, grassroots way, which has been OpenEvidence’s strategy, you marry the best of what we all admire the Mayo Clinic and the Cleveland Clinic for, which is a team of experts, and you marry that with the ability to serve and help the rest of the country,” he said.