Daniel Nadler started OpenEvidence to help physicians sort through a deluge of medical research. Now, he’s raised $210 million at a $3.5 billion valuation.
For doctors trying to stay abreast of the latest medical breakthroughs, reviewing the latest research is like being shot in the face with a water cannon. A new paper is published every 30 seconds. Trying to comb through it all to come up with a diagnosis or treatment plan that reflects the best current options while seeing 20 patients a day is a near-impossible task.
“We talk about the golden age of biotechnology where there are new drugs and better drugs developed all the time. But it is like the dark ages for physicians because of burnout,” Daniel Nadler, cofounder and CEO of OpenEvidence, told Forbes. “There is this enormous firehose of information they need to stay on top of, and the human brain is limited in its capacity to read millions of studies.”
Mauricio Candela for Forbes
So Nadler, 42, a Harvard Ph.D. who sold his previous company for $550 million back in 2018, set out to solve the problem with artificial intelligence. Now, the startup’s proprietary algorithms search millions of peer-reviewed publications, including in top journals like the New England Journal of Medicine and the Journal of the American Medical Association, to help doctors find the best answers fast, with full citations to papers so doctors can read more for themselves. The software is free for verified doctors to use and makes money through advertising–much like Google does.
“I think OpenEvidence looks like it’s going to be for healthcare what Google was for the Internet,” said Kleiner Perkins billionaire chairman John Doerr, who invested in the company personally as well as through his firm, adding, “It’s the free-for-physician model that’s the magic here.”
Since its founding in 2022, Miami-based OpenEvidence has signed up 40% of doctors in the United States, or more than 430,000, and is adding new ones at a current rate of 65,000 per month. Its revenue from advertising is now coming in at an annualized rate estimated at $50 million. That’s not huge, but thanks to the software’s rapid adoption, investors are betting big: OpenEvidence has raised $210 million led by GV (Google’s venture arm) and Kleiner Perkins at a valuation of $3.5 billion, up from $1 billion at its last financing in February, Nadler told Forbes. Other storied VC firms like Coatue, Conviction and Thrive Capital also invested.
The new investment makes Nadler, who owns roughly 60% of the company, a billionaire, with a net worth that Forbes estimates at $2.3 billion. Cofounder Zack Ziegler, the company’s 30-year-old chief technology officer, owns some 10% of the business, worth about $350 million. Nadler was able to hold on to such a large stake by being its first seed investor, putting in some $10 million of his own money before raising any VC funding.
“One of the great things about being a second-time entrepreneur is, I’m not an idiot,” Nadler said. “I think the second thing is going to be bigger than the first so maybe the first $10 million should come from me. That’s by far the smartest financial decision I made in my life…..I wanted to bet on myself.”
The problem OpenEvidence is addressing is enormous, and one that’s only getting bigger. Medical literature is proliferating at a meteoric rate—doubling in size every five years—as new treatment options like gene therapies are developed and scientists learn more about how different diseases and drugs may interact with each other. Sorting through all of it is a herculean task: some papers are excellent, some are bad and many more are outdated. (With AI being used to publish and review research papers, the problem has only gotten worse.) Meanwhile, physicians in the United States are increasingly strapped for time, given a growing shortage of medical professionals—creating an opportunity for startups to build technology that can help provide better care and relieve pressure on doctors.
“When everyone was scrambling to come out of crypto, I was like, ‘I am just going to run circles around all of you.’”
OpenEvidence isn’t the first company to try to make sense of the overload of medical publications; Wolters Kluwer’s UpToDate has been around for decades and has recently been incorporating AI, along with advice from experts, to do the same thing. But it is the first to build software that integrates AI from the start to make it easier for doctors to find answers to pressing clinical questions and to do so far more accurately than ChatGPT.
Doctors now use OpenEvidence on some 8.5 million consultations a month. Because the tool isn’t considered diagnostic, it doesn’t need FDA approval as algorithms used to detect strokes or sepsis in patients do. And since doctors can download it or use it online for free, it can bypass the lengthy and bureaucratic procurement process with hospitals or large group practices. That’s helped the company sign up doctors at an ever-faster clip.
Dr. Susan Wolver, an internist in Richmond, Va., has become a true believer, using OpenEvidence to write prior authorization letters and look up details of drugs. Most dramatically, while she was on a domestic flight recently, an immunocompromised passenger nearly fainted in the bathroom. Wolver turned to OpenEvidence to figure out the patient’s immune-system risks and come up with a treatment plan on the spot.
“I don’t think a day goes by when I don’t use it,” she said.
Nadler grew up in Toronto where his parents were part of the great wave of post-war Eastern European immigrants–his father from Romania and his mother from Poland. “My grandfather was in Auschwitz and survived,” he said. “After the second world war, my grandfather wanted to come to America, and America wasn’t letting people in so they made their way to Canada.”
As a kid, Nadler was competitive to a fault, dabbling in memory games to see whether he could recite more pages of a Hamlet soliloquy than a friend. “I was a total nerd,” he said. A Mensa member, he found school boring, and after getting a bachelor’s degree at the University of Toronto, applied to Harvard for grad school, hoping for a greater challenge. Once there, he got a Ph.D in political economy, writing his thesis on the pricing mechanisms of credit derivatives. He also studied poetry under Pulitzer Prize-winner Jorie Graham, launched an app called Sigmund that could be programmed to speak specific words during sleep to influence a user’s dreams and served as a visiting scholar at the Federal Reserve.
Nadler was working on his Ph.D., and making just $23,500 a year as a grad student, when he got the idea for his first company, Kensho. At the Fed, he’d been stunned to learn that its regulators relied on rudimentary Excel spreadsheets to make critical assessments. So he teamed up with programmer Peter Kruskall to build algorithms to make financial analysis as easy as a search on Google. When Kensho launched its text-based chatbot, Warren (as in Buffett) in 2012, artificial intelligence was still the province of academics, not the red hot center of the startup world that it is today. “No one was talking about AI in 2012. You are talking about 10 years before ChatGPT,” he said.
The idea worked, and when S&P bought Kensho, paying $700 million, including retention bonuses, it became the largest AI deal in history. Nadler, who owned 20%, was suddenly rich. “For second-time founders, that hubris is [often] gone,” said GV partner Sangeen Zeb. “Daniel still has that hubris.”
In 2021, he teamed up with Ziegler, who was working on a Ph.D. in machine learning at Harvard, but really just wanted to build stuff. The two had a hunch that the AI technology that had helped make traders smarter by finding patterns in large swathes of data could also aid doctors–with even greater impact. Both were also motivated by personal experience. Nadler’s grandfather had died due to a medical error, while Ziegler had watched his brother-in-law, then 22, go through treatment for leukemia. (He’s in remission now.) “It was really eye-opening to me,” Ziegler said. “There’s this enormous amount of complexity, but the way doctors are accessing it is literally leafing through a textbook.”
“I think OpenEvidence looks like it’s going to be for healthcare what Google was for the Internet.”
Venture capitalist Jim Breyer, who had invested in Kensho, spent four hours talking with Nadler about his idea for OpenEvidence and became one of its first outside investors (along with billionaire investor Ken Moelis) in 2022. Breyer, who’d famously backed Mark Zuckerberg in 2005, considers Nadler among a rarified group of founders. “Daniel is an extraordinary entrepreneur,” he said. “The initial insight of applying AI to medical journals was simply brilliant.”
In early 2023, OpenEvidence joined the Mayo Clinic’s prestigious accelerator for health tech startups. The program allows startups to refine their ideas—and their technology—at the hospital that, as Nadler noted in a 2023 video for the program, “has the largest and highest quality dataset in healthcare.” By this point, AI was booming. Nadler’s decade in the field quickly began to pay off. “When everyone was scrambling to come out of crypto, I was like, ‘I am just going to run circles around all of you,’” he said.
Still, this is a tough business and there are questions about whether AI-based search will always give the best answer. Nadler argues that by relying on the “gold standards of medical knowledge,” many of which aren’t available on the open internet beyond abstracts—including JAMA and the New England Journal of Medicine—the startup’s search ranking models are able to extract reliable and relevant information about a rare disease or a drug’s side effects, while keeping hallucinations (AI’s tendency to fabricate facts) at a minimum. “AI is garbage in, garbage out, gold in, gold out,” Nadler said, adding, “Not everything is about coming up with a super nerdy algorithm.”
Dr. Stephen Krieger, a multiple sclerosis specialist at Mount Sinai in New York, heard about OpenEvidence from a resident this past weekend when he was doing rounds in the hospital. He needed to find out what antibiotic to use for a neurological infection in someone allergic to penicillin, something outside his own clinical expertise. Before relying on it, he tested its accuracy by asking OpenEvidence about his own research on MS (and also confirmed its answer with his infectious disease colleagues). Not only did it correctly summarize his research, it properly noted constraints that had yet to be published. “The idea that it offered to tell me the limitations of my own work and I agreed with it I thought was kind of great,” he said.
“One of the great things about being a second-time entrepreneur is, I’m not an idiot.”
But Daniel Byrne, a lecturer at Johns Hopkins Bloomberg School of Public Health and author of the book Artificial Intelligence for Improved Patient Outcomes, said it’s not so simple. “What I found that most people misunderstand is that up to half the medical literature is wrong,” he said, noting that papers will often be published about scientific debates or clinical studies that may not ultimately pan out. “Having a reference is a step in the right direction, but it’s just not enough,” Byrne said.
Dr. Travis Zack, OpenEvidence’s medical director, says that while there will be errors from any AI system, there should be far fewer than with doctors making judgment calls for 20 patients a day without easily consulting the available literature. “What OpenEvidence does is allow physicians to not have to trust their gut,” he said.
“AI is garbage in, garbage out, gold in, gold out.”
It also remains to be seen how successful OpenEvidence’s ad model will be. Pharmaceutical companies are big spenders, and now they have an opportunity to get detailed information about their drugs in front of doctors who are likely to use them. Thanks to sponsored answers, the company is able to keep the tool free for doctors, helping attract more clinicians and allowing it to tweak its algorithm (and improve search results) based on their feedback. That creates what Nadler calls a “fantasy flywheel,” in which having more users makes the product better, which draws more users, ad infinitum.
But despite healthcare and pharma ad spending amounting to some $30 billion in 2024, building an ad-based business is unusual in healthtech, where most software is sold on a subscription basis. “People hate on advertising,” Nadler said. “I don’t know why—I love advertising.” But even he notes that the company currently has a far greater potential inventory of ads, more than $350 million, than it has sold to date. “Google spent time getting people comfortable with the model, and that’s what we are doing.”
Dr. Aneesh Singhal, vice chair of Massachusetts General Hospital’s department of neurology and director of the hospital’s stroke center, downloaded OpenEvidence a year ago, after reading about it in a mass email sent out to the hospital system. Since then he’s noticed the tool gain popularity among his residents and surgeon colleagues. “Everybody seems to be using it,” he said.
He wanted to search for the latest studies on stroke in adults—a daunting task that would otherwise take hours of rummaging through PubMed and online textbooks. The tool proved far better than a generic chatbot like ChatGPT, suggesting follow-up questions to ask about a patient’s medical history and tests that should be conducted, he said. “ChatGPT stops short in that it just gives you the direct answer,” Singhal says.
OpenEvidence’s momentum so far has been staggering as it signs up doctors at an ever-faster clip, a key metric that investor Breyer wants to see. “Getting the weekly and monthly updates gives me enormous confidence that Daniel continues to knock it out of the park,” he said.
Now it’s pushing into using so-called reasoning models, which think through a task in steps, a tactic researchers have found makes AI answers better and more robust. This month, the startup launched a new feature called DeepConsult, which uses this technique to connect the dots between different studies and carry out advanced research on a particular topic. “It allows a physician to essentially have a team of M.D. Ph.D.s that can go off while the physician is doing other things and do that enormously deep amount of research,” cofounder Ziegler said.
And while OpenEvidence’s tech could be used in a similar way for other scientific fields, Nadler isn’t focused on expansion there yet: he wants to stick to healthcare, both in the U.S. and internationally, especially in countries where access to quality care is limited. Across the industry, there is now a mosaic of AI-powered technology, from notetakers for doctors to clinical diagnostic tools. Pair that with a patient’s lab results and data from medical devices like blood glucose monitors, and there’s an opportunity to bring all of that information together, in one place.
Coatue cofounder Thomas Laffont, who is invested in OpenEvidence, sees the startup someday becoming the hub where all these tools might converge. “You can squint to a world where OpenEvidence becomes the tool through which all that diagnosis is happening,” he said.
More from Forbes
Source link
