The most instructive company on Doug Proctor’s résumé is the one that failed. Insula had been formed to commercialize an insulation material developed at a United States national laboratory. The proposition had everything an ambitious founder might wish to put on a pitch deck: large energy stakes, plausible climate benefit, attractive manufacturing characteristics and defensible intellectual property. Proctor assembled a team to find out whether the central promise was real. They found out. It was not.
The team invalidated the core technology. Then it returned capital to investors and wound the company down. No heroic pivot was retrofitted onto the ruins. No vocabulary was invented to make closure resemble momentum. The evidence arrived, and Proctor allowed it to win.
That decision is an unusually clear window into the operator who now serves as co-founder and chief operating officer of Candid Health. The company works on medical billing, a field where wishful thinking has enjoyed a long and expensive residency. Its object is to automate the transaction machinery between care providers and insurers: the payer rules, rejected claims, corrections, follow-up and posting that determine whether a provider gets paid. This is plumbing, if plumbing had more than a thousand rulebooks and every pipe could dispute the invoice.
“Doug is fundamentally motivated by learning and by building effective teams which meaningfully impact hard and worthy problems.”Proctor’s own description of his north star
Proctor’s public account of his career is not organized around a beloved sector. It is organized around difficult systems. He grew up in Washington, D.C., attended Sidwell Friends School, lived in Beijing for two years before college, and graduated from Georgetown University’s School of Foreign Service. He lists professional working proficiency in Mandarin. The young internationalist eventually went to work on software.
Learning the shape of consequential data
In 2015, Proctor joined Palantir. He began as chief of staff to the head of product and engineering while the company was resetting its product stack, work that led to Palantir Foundry. He then spent more than four years building and leading product and business teams in defense. According to his account, those teams worked on software supporting global intelligence and operational activities at U.S. Special Operations Command, the Army’s replacement of DCGS-A CD2 with TITAN, and Army Vantage, a data platform used across levels of the service.
His future Candid co-founder Nick Perry was at Palantir too, leading healthcare work. Their domains looked different. Their diagnosis was similar. Both fields had fragmented data, legacy systems, local workarounds and decisions too important to leave to a cheerful demo. Medical billing, they would later argue, was another information-integration problem hiding in institutional clothing.
Then came Insula and its beautifully inconvenient result. The experience could be read as an interruption between two software companies. It is more useful as a bridge. Defense systems had taught Proctor the cost of ambiguity. Insula taught him that a disciplined team must be able to disprove its reason for being. Candid would need both lessons.
A comma can hold up the money
American healthcare billing is not merely complicated. It is granular. A provider sends a claim; an insurer judges it against a private, changing collection of rules. A wrong code, missing field or tiny formatting defect can trigger a rejection. Someone then has to discover the reason, correct the claim and try again. Multiply the ritual across specialties, locations and payers, and “revenue cycle management” stops sounding like back-office jargon. It becomes the operating system of whether care businesses can collect what they are owed.
Candid’s founders learned the billing process by hand before attempting to automate it. That apprenticeship matters. Their ambition was not to attach a newer interface to the same burden. It was to encode the logic, normalize the data and prevent avoidable errors upstream. Proctor’s responsibility at Candid spans the product, customer and people teams, and he sits on the board. In other words, the logic of the machine and the habits of the organization meet on his desk.
There is a strategic consequence to learning the work at that depth. Candid can replace a billing platform, but it can also sit above parts of an existing system and orchestrate them. That makes the product less like a single clever application and more like an infrastructure layer. In a 2024 conversation, Proctor and Perry compared the long-term opportunity to Stripe or Shopify: software that handles enough underlying complexity that a new business can begin without first assembling an administrative department. The comparison is ambitious, but the mechanism is concrete. Every rule the platform can apply correctly and every exception it can expose clearly is one less private ritual a billing team must preserve in a spreadsheet or in somebody’s memory.
The results have attracted capital. In July 2026, Candid announced a $120 million Series D led by Sixth Street Growth, with Oak HC/FT, 8VC and Y Combinator participating. The round followed a $52.5 million Series C and a $29 million Series B. Annual recurring revenue had grown 190 percent over the previous year. The new round tripled the company’s 2025 valuation and took total funding past $219 million.
Money is the easy part of the story to chart. The more distinctive fact is how little the company had previously spent on marketing. Perry said growth came largely through word of mouth. In software sold to people responsible for collecting revenue, enthusiasm is nice; trust renews the contract.
Capital followed the operating proof
Automation must earn its verbs
At ViVE in 2026, Proctor offered a useful rebuke to the season’s AI exuberance. Deploying large language models, machine learning or automation, he said, is something a company has to “earn your right to.” Candid’s approach had been methodical: use technology internally where the impact could be observed, build trust in it, determine where it created value, and only then deploy it with customers.
This is a less cinematic theory of innovation. It also sounds like the Insula lesson translated into software. A technology does not deserve scale because its promise is large. It deserves scale when sustained contact with reality has made it dependable.
“Constraints really force a degree of creativity.”Doug Proctor at ViVE 2026
The same principle appears in Proctor’s account of Candid’s culture. Its leading value, he has written, is to “relentlessly orient around customer outcomes.” During one complex customer launch, a legacy component in the customer’s stack caused a major interruption. Candid employees spread across dozens of physical and remote locations to help hundreds of new users resume important workflows. The customer’s note afterward praised the team for adapting under a tight timetable and solving an unusually complicated go-live.
This anecdote is not tidy. The interruption still happened. People had to fan out and fix things. Yet it says more about an operating culture than a plaque of values ever could. The relevant question was not who owned the fault. It was whether the users could work.
The glamour of competent plumbing
A Fortune reporter’s first video call with Proctor acquired one surreal grace note. He turned his laptop toward a herd of wild horses grazing behind him in upstate New York. It was an excellent view from a conversation about billing administration. Proctor and Perry, the story observed, had deliberately built outside the usual Bay Area theater, favoring glasses and black T-shirts. Even founders cannot fully escape a uniform.
Proctor is based in Denver, far from Candid’s San Francisco headquarters but close to the company’s operating work. His stated ambition is not personal celebrity. He talks about learning, teams and worthy problems. For Candid, the five-year vision is a healthcare payments system that costs less and works more efficiently. The company’s sharper commercial proposition is that providers using its infrastructure should collect more of what they are owed while spending less human effort to do it.
There is a moral temptation in difficult industries to confuse complexity with permanence. Billing has always been awful, the thinking goes, and therefore awfulness must be intrinsic. Proctor’s career argues for a more demanding view. Complexity should be studied until its rules become visible. Technology should be tested until it earns trust. Teams should be built to serve outcomes rather than defend boundaries. And when the premise fails, the honest answer may be to stop.
The $120 million round says investors believe Candid has found a premise worth scaling. The closed company in Proctor’s past says he knows the alternative. Taken together, they form a rather good operating creed: pursue the hard problem, but keep reality in charge.