The surgeon who looked at a trauma patient and realized medicine was flying blind without data. He built two companies to fix that - and isn't done yet.
Picture an operating room in Los Angeles. A car crash victim arrives with multiple organ injuries. The trauma surgeon - Dan Riskin - knows the immediate moves. What he doesn't have is any way to look at what happened to similar patients before. No aggregated outcomes. No comparable cohort. Just instinct, training, and incomplete information.
That moment stuck. Riskin had spent his career moving between disciplines most people treat as separate worlds - medicine and software, clinical care and policy, surgery and bioinformatics. He saw a gap that wasn't philosophical but practical: medicine accumulates vast data about what happens to real patients, and almost none of it gets organized well enough to be useful.
The promise of "real-world evidence" - learning from routinely collected patient data rather than only from tightly controlled clinical trials - was everywhere by the 2010s. The problem was that the data was a mess. Electronic health records contain the truth about patients, but that truth is buried in unstructured notes, inconsistent coding, and fragmented systems. The industry was building conclusions on shaky foundations.
Riskin founded Verantos in 2015 with a deceptively simple premise: if real-world evidence is going to guide drug approvals, clinical decisions, and reimbursement policies, then the underlying data needs to be trustworthy. Not just big. Trustworthy.
Eighty percent of clinical data lives in the unstructured narrative sections of physician notes - the free-text where doctors actually describe what they observed. Riskin built Verantos' AI to read those notes. Without it, the claim to understand a patient population is, at best, partial.
Verantos achieves data accuracy above 85% in the studies it performs. The industry norm sits at 30 to 40%. That gap - which most organizations never measure - is the difference between evidence that influences care and evidence that misleads it.
Verantos vs. Industry Standard
Dan Riskin grew up in the suburbs of Los Angeles, and by five years old he was already writing software. Not playing games - writing them. By twelve, he was selling software. By fifteen, he had won the Bausch & Lomb Outstanding Scientist Award. At sixteen, he walked into UCLA as a Regent's Scholar.
This is not a career arc of someone who found their calling late. Riskin arrived already pointed. The technology instinct was there before the medical one, which is what makes his eventual combination of the two feel less like a pivot and more like convergence.
He earned an MD from Boston University, completed a surgery residency at UCLA, and a fellowship in critical care and acute care surgery at Stanford. Then he added an MBA from MIT, focused on bioinformatics. He holds four board certifications across surgery, critical care, palliative care, and clinical informatics. Each credential doesn't just stack - it connects to the others.
While still a student, Riskin designed a wound closure device. It received U.S. patents and FDA clearance. The company he co-founded around it - Wadsworth Medical Technologies - was later acquired. The device ended up used on hundreds of thousands of patients. This was before his surgical training was complete.
Riskin sat down with Taiwanese government officials for a dinner about healthcare strategy. He'd informed the hosts beforehand that he was vegan. They served twelve separate full-course meals anyway. Cultural custom, it turned out, meant honoring the guest rather than the request. By course six he was struggling. He finished all twelve.
Before Verantos, there was Health Fidelity. Riskin founded it in 2011 with a version of the same conviction: if you apply AI to clinical data, you can extract meaningful signals that improve care. Health Fidelity focused on quality measurement and risk adjustment, reading unstructured EHR data to give payers and providers more accurate pictures of their patient populations.
Edifecs acquired it for more than $150 million. Most entrepreneurs call that a finish line. Riskin treated it as a calibration. He saw what worked in the clinical data space, what was still unsolved, and where the incentives were misaligned. The second company - Verantos - went deeper on the hardest part: making real-world evidence reliable enough to guide regulatory decisions.
Verantos' customers include major pharmaceutical companies, academic medical centers, and health systems. The platform integrates data from electronic health records, medical claims, and disease registries. It then validates that data against source documents - the kind of rigorous verification that most RWE platforms skip because it's expensive and slow.
Riskin's position is that skipping it is also expensive - just in a different currency. Low-quality evidence produces bad conclusions. Bad conclusions affect treatment standards. Treatment standards affect patients.
"The goal isn't just to measure data reliability; it's to optimize it and to do so using peer-reviewed, published methods so metrics can't be gamed."Dan Riskin — Verantos
"If that evidence is low quality, we will see a harm in the standard of care."Dan Riskin — on real-world evidence standards
"Know when the customer doesn't know. There's an opposing view for every axiom. Sometimes a customer doesn't know what they want until it's shown to them."Dan Riskin — on entrepreneurship
"In the studies we perform, data accuracy is unparalleled; greater than 85 percent compared to industry norms of 30 to 40 percent."Dan Riskin — Verantos performance data
"Rich data sources include the narrative sections of doctor reports, representing 80% of the EHRs."Dan Riskin — on where clinical truth lives
"If you look at their most recent framework in RWE, data validity - it's trying to get at what is believable."Dan Riskin — on FDA's RWE framework
In 2014, Riskin testified before the U.S. House of Representatives about the 21st Century Cures Initiative, legislation designed to accelerate medical product development using real-world evidence. He then met with the Congressional retreat to help refine the actual language of the bill.
He has served on healthcare advisory committees under both the Obama and Biden administrations, and Congress appointed him to the Health Information Technology Advisory Committee (HITAC), where his term runs 2024-2026. He has presented at the FDA, CMS, NASA, DARPA, NSF, and NIH - an unusual roster for a physician-entrepreneur.
Beyond the U.S., Riskin has advised governments in Europe and Asia on digital health policy. The problems he works on - how to make clinical data reliable, how to set standards for real-world evidence, how to regulate AI in healthcare - do not stop at national borders.
His mentor David Morgenthaler, an MIT board member, pointed him toward Silicon Valley early. Surgical innovators Tom Fogarty and Tom Krummel at Stanford shaped his approach to bringing technology into clinical care. Riskin's career is, in part, the product of a network that saw across disciplines before he did.
Started coding at age 5 - the same year IBM launched its first PC.
Entered college at 16. His peers were still in high school.
A wound closure device he designed as a student ended up used on hundreds of thousands of patients worldwide.
Has spoken at NASA and DARPA - not typical venues for a physician-entrepreneur.
His favorite quote belongs to William Gibson: "The future is already here, it's just not evenly distributed."
The company he built before Verantos - Health Fidelity - sold for more than $150 million.
Advised governments in Europe and Asia on digital health, in addition to two U.S. administrations.
One of very few people board-certified in both surgery and clinical informatics simultaneously.
In this interview with the Health Business Group, Riskin explains why first-generation real-world evidence often fails - and what a rigorous alternative looks like.
Key themes: the difference between data volume and data validity, why unstructured EHR text matters, and what "believable" evidence actually means to regulators.