There is a canoe somewhere in the origin story of InStride Health. James McElhiney remembers being in it with three future co-founders - clinicians Mona Potter and Kathryn Boger, and executive John Voith - talking through a company that did not yet exist. Their proposition was ambitious and specific: organize specialized care around a coordinated team, make it work virtually, and have insurance pay for it. The image is almost suspiciously tidy. Four people, one narrow boat, no room for anyone to paddle in a private direction.
McElhiney, usually called Jaimie by colleagues, had already spent decades learning the organizational version of that lesson. A technical system can be ingenious and still fail if its users are pulling against it. A company can have capital, talent, and urgency and still go in circles. The code matters. So do the sequence, the handoffs, and the quiet agreement about where everyone is trying to go.
His career looks eclectic from a distance. He helped start Gazelle, which gave consumers a price for used electronics. He became the founding chief technology officer at Iora Health, which rebuilt primary care around teams. He co-founded Corvus Insurance, where data informed commercial risk. Then came InStride, a virtual behavioral health company for young people and families. Phones, clinics, cyber policies, care plans: a peculiar collection until you notice the repeated engineering problem beneath it. Each company tried to make a complicated service legible enough to use.
Four chapters, one recurring craft
The patent trail
Long before the canoe, McElhiney was an electrical-engineering student at Worcester Polytechnic Institute. He graduated in 1986 and entered a working world of test equipment, enterprise machines, and networks that still had to be explained to normal people. His early résumé runs through Teradyne, Unisys, and software consulting. One patent application filed in 1997 named him on a system that captured analog audio or video, encoded it, attached identifying information, stored it, and made it searchable online. The language belongs to another internet age. The underlying interest does not.
A second cluster of patent applications arrived in 2008, around the time Gazelle launched its consumer site. They covered dynamic pricing, trade-ins, product intelligence, marketplace selection, and the syndication of listings. Put less ceremoniously: how do you look at an old gadget, decide what it is worth, take it in, and find the sensible place for it to go next?
Gazelle had turned the junk drawer into a supply chain. McElhiney co-founded the business with Israel Ganot and Rousseau Aurelien in 2006. Consumers received offers for phones, tablets, and computers that might otherwise have collected dust. The company later processed millions of devices and was acquired by Outerwall in 2015. For McElhiney, though, the revealing artifact is the machinery: pricing rules, product profiles, transaction workflows, and market connections. The customer saw an offer. The system had to see the entire journey.
Software meets the examination room
In 2011, McElhiney joined Iora Health as founding CTO. The company’s care model relied on physicians, nurses, and health coaches working as a unit. Conventional medical records were not designed around that cast of characters, so his group built a platform of its own. It became known as Chirp.
John G. Norman, whom McElhiney recruited as chief software architect, later described the assignment with a warning they heard at the outset: building the system would be a “land war in Asia.” The metaphor was gloomy; the work was granular. Labs, pharmacy, referrals, claims, clinical tasks, and conversations all had to meet in one place without making the people delivering care serve the software.
The enduring technical decision was to design around the care team, not around the software department.Iora Health, 2011-2016
Norman credited McElhiney with patient guidance, careful build-versus-buy decisions, and a steady hand. He also pointed to a less glamorous metric: over several years, only one engineer left the group. Staff continuity does not fit neatly on a product roadmap, but it changes what a team can remember. It lets technical judgment accumulate instead of resetting every hiring cycle.
That combination - architecture and continuity - helps explain McElhiney’s next move. Corvus Insurance, co-founded in 2017 with insurance executive Phil Edmundson, used data to help underwrite and monitor commercial risk. The domain changed again, from clinics to policies. The structural question remained familiar: how can a system collect signals, make them intelligible, and deliver a useful decision to the person responsible for acting?
There is a temptation to read each move as a sharp turn. In fact, the distances are shorter than they appear. Gazelle needed confidence between a stranger, an unseen device, and an automated offer. Iora needed confidence between a clinician, a colleague, and a shared record. Corvus needed confidence between underwriters, brokers, customers, and a stream of external data. None could succeed by presenting more information merely because software made more information available. The job was to select, sequence, and expose enough of it for the next person to act.
That is closer to editorial work than technological spectacle. A useful platform decides what belongs in the foreground, what can wait, and what should never become the user’s burden. It also accepts that the decisive knowledge may sit outside engineering. Clinicians understand the encounter. Underwriters understand the judgment. Operations teams understand the exception that ruins an otherwise lovely flowchart. McElhiney’s public record offers no manifesto on this point, but the companies themselves form an argument: technical leadership in a service business begins by treating domain expertise as part of the architecture.
Back to the boat
McElhiney left Corvus in 2020 and began the work that became InStride. Potter and Boger had developed an intensive clinical program at McLean Hospital. They understood the treatment model and the access problem. Voith brought company-building experience. McElhiney brought the pattern language of technical systems that live inside services.
When InStride formally launched in 2022, he wrote that the company had been in progress for more than two years. Its model combined therapists, psychiatrists, and coaches, with mobile apps supporting patients and families between appointments. The technology was not offered as a substitute for the care team. It was the connective tissue: messaging, video, measurement, and coordination around work that also happened in homes, schools, and ordinary daily settings.
The company announced $26 million in backing at launch. In 2024, it raised a $30 million Series B led by General Catalyst. McElhiney returned publicly to the canoe memory, writing with unusual economy: “We’re doing it!” By then, InStride was operating across several Northeastern states. He moved from the CTO role into an advisor and co-founder position that spring, leaving the day-to-day technology post while remaining connected to the company.
In June 2026, InStride announced another $30 million round. The company said it was operating in 17 states and had served more than 5,000 patients. Those are company figures, not a solo achievement, which is precisely the point. McElhiney’s public language repeatedly spreads credit across clinicians, operators, investors, and builders. The technology leader appears most comfortable describing a crew.
A founder’s architecture is partly code, partly judgment, and partly the conditions under which other people can keep doing good work.The pattern across four companies
The latest startup has a drummer
There is one more company in McElhiney’s telling, though the incorporation papers may be imaginary. In 2024 he announced his “latest startup,” a band called Yurrn. Lilly Darr sings and plays flute and saxophone, William Crawford plays drums, and McElhiney plays guitar. Their first single, “Zombie,” arrived that October, with all three credited as songwriters.
It is a good joke because a band resembles a startup just enough to be dangerous. There is an idea, a small group, an audience that owes you nothing, and no amount of individual cleverness can rescue bad timing. It also reveals a preference already visible in his engineering career. McElhiney is drawn to ensemble problems.
The romantic version of technology celebrates the virtuoso. McElhiney’s record suggests something more durable and more social. At Gazelle, the customer’s simple quote depended on a chain of decisions. At Iora, clinical work depended on shared context. At Corvus, risk information had to reach brokers and policyholders in useful form. At InStride, software supported a coordinated care team. In Yurrn, the guitar is only interesting because someone else keeps time.
A career can look like a series of industries, or it can look like the gradual refinement of a question. McElhiney’s question has followed him from encoded media to used phones, from examination rooms to insurance, from a canoe to a mobile app: what must be connected for a group of people to move together? The answer is never only software. That is why the software matters.