The room went quiet. John Bertrand had just finished laying out a five-year vision for Digital Diagnostics: the long-range goals, the grand ideas, the version of the company that might exist if everything worked. Then a member of the leadership team punctured the altitude with six practical words: “Okay, but what do we do next week?”
Bertrand has told the story on himself. The vision came naturally. The sequence did not. His team did not need another horizon; it needed the next tactical milestone. The exchange became a small operating parable, and perhaps the cleanest way to understand his career. Bertrand works at the seam where an ambitious technology meets Monday morning.
That seam is rarely glamorous. Scientific progress must become a product. The product must fit a real workflow. The workflow must be understood by a buyer, accepted by a professional, cleared by a regulator, paid for by a system, and supported after installation. Any one of those steps can turn a breakthrough into a museum piece. Bertrand's public record is largely about making the steps connect.
An apprenticeship in the plumbing
Bertrand arrived at Epic Systems in 2006 after earning a business management degree from Purdue University. He stayed until 2019, moving through work that spanned product, professional services, customer success, sales, and business development. It was less a single job than a tour through the machinery of healthcare software.
The background mattered. Bertrand has said he always assumed he would work in healthcare. Both of his parents were in the field, and new technology was dinner-table conversation. At Epic, that general attraction became specific: software is never merely software once it enters an institution. It changes tasks, responsibility, timing, and the movement of information.
Early in his career, he realized that a design mistake could cause data to be misplaced or overlooked. The consequence would not be an ugly screen or an annoyed customer. Someone could be harmed. He carried that gravity into his later arguments about AI: validation is not paperwork around the product. In consequential systems, validation is part of the product.
“The guiding principle here should be to first understand the problem that needs solving.”John Bertrand, 2024
After Epic, Bertrand spent time at 8VC as an Executive in Residence and adviser, looking at artificial intelligence and computer vision in healthcare. From the investing side, he could see impressive work in laboratories and research institutions, plus the long distance between impressive and adopted. The science might be ready before the environment was.
The chapter after “first”
In December 2019, Bertrand became CEO of IDx, the company later renamed Digital Diagnostics. Its founder, Michael Abramoff, had already led the creation of IDx-DR, later called LumineticsCore, and the system had received FDA authorization in 2018. Bertrand is careful about the chronology. He has credited that milestone to Abramoff and the team before his arrival.
His assignment was the chapter after “first.” Digital Diagnostics needed to move from a research-oriented company toward a commercial operation. The practical work included adoption, deployment, partnerships, and payment. In 2022, the board gave Bertrand and President and COO Seth Rainford co-founder designations, recognizing their impact on the company's expansion. That August, Digital Diagnostics announced a $75 million Series B round led by KKR.
Bertrand points to reimbursement for the company's autonomous AI product as the moment when he felt like “a little footnote” in the history of the industry. The phrase is modest, but the operating insight is sharp. Clearance answers whether a product may enter the market. Reimbursement helps answer whether the market can routinely use it. An invention without a payment path is still waiting at the door.
This is where the familiar story of innovation becomes a little less cinematic and much more consequential. A product can pass a technical test yet fail an institutional one. Somebody must decide who performs the task, where the result appears, how an exception is handled, which code is submitted, and what happens when the ordinary day becomes an unusual day. Each answer creates another dependency. The operator's work is to turn that web into a route people can follow.
Bertrand's language around this work is notably short on conquest. He talks about working within the healthcare system, not routing around it. He treats stakeholder agreement as an engineering constraint: if the people responsible for safety, payment, and daily use are absent from development, the missing context will surface later as resistance. Consultation at the beginning is cheaper than repair at the end.
Working with Rainford, Bertrand had to explain something genuinely new through concepts that institutions already understood. Innovation needed an analogy. The team had to preserve what was different while making the adoption decision legible. This was translation as company building.
A translator with a wide listening radius
Bertrand's leadership language keeps returning to listening. A healthcare product touches what he calls a “whole rainbow of stakeholders,” from the front desk to the clinician, plus developers, regulators, payers, and customers. Each group encounters a different product because each group carries a different risk.
His stated role is to collect those perspectives, synthesize them, and offer clarity to people who were not in every conversation. This is an operator's version of product management at organizational scale. The CEO is not the person who knows every answer. The CEO builds a system in which the right information can reach the right decision.
Bertrand is unusually direct about the corollary: sometimes the company rides on a decision made by somebody other than the CEO. In regulated environments, authority without expertise is a hazard. He has described relying on specialists closest to reimbursement or regulation, even when he does not possess every detail himself. Trust becomes a design choice.
That choice also changes the emotional geometry of a team. Delegation can sound like a leader stepping away. Bertrand frames it as the opposite: getting close enough to the work to know whose judgment deserves weight. The leader keeps the common mission visible, makes the decision boundaries clear, and allows expertise to travel without waiting for every answer to climb the org chart. Speed comes from a better map of responsibility, not simply from urgency.
“As a leader you have to get comfortable with the fact that sometimes the company's going to ride on somebody who is not me making a call.”John Bertrand, 2022
The same instinct shapes his view of the market. Bertrand is wary of technology in search of a problem, and of AI presented as a margin-enhancing feature without demonstrated value. His test has four parts: does it improve outcomes, reduce costs, expand access, or improve equity? The list forces a builder to name the benefit before admiring the mechanism.
He is equally skeptical of the mythology that a company can simply force its way through healthcare. His preferred verbs are collaborate, validate, integrate. Regulators and domain specialists are not obstacles outside the work. They are participants in the work. The product earns trust by showing its reasoning, fitting established standards, and respecting the people responsible for its use.
Begin with an expensive, frequent, clearly defined problem. Let the technology remain a tool.
Build validation, workflow, and payment into the product plan before the launch slide.
Give specialists real decision rights when their context is better than yours.
Pair the five-year story with a milestone the team can complete next week.
Inning two
Bertrand compares the current state of AI to the second inning of a baseball game. The data has become more organized. The science has matured. Early applications are showing real use. But most of the game remains, and the metaphor quietly refuses both panic and premature victory.
His own career now moves into another part of the system. He finished his Digital Diagnostics CEO tenure in December 2025. In 2026, he became Chief Innovation Officer at PartsSource, where his remit connects data and AI with the orchestration of diagnostic and medical equipment. In April, he also joined the board of OraSure Technologies as an independent director and member of its nominating and governance committee.
The move is consistent with the rest of the story. Bertrand keeps drifting toward the point where technology encounters installed reality. At Epic, that reality was the clinical software environment. At Digital Diagnostics, it was the route from validated system to commercial platform. At PartsSource, it is the equipment, uptime, data, and operational capacity already inside healthcare organizations. The unit of attention widens from an algorithm to the assets around it.
There is no clean ending to an operating story because operations are recurrence. A system works today and must work again tomorrow, for the next customer, inside the next workflow, under the next constraint. Bertrand's useful idea is that progress happens when a large mission is continuously translated into small, credible commitments.
Hold the five-year picture. Listen to the person who sees a piece you missed. Prove the claim. Respect the system. Then answer the question waiting at the end of the presentation: what, exactly, do we do next week?