The useful detour
Mark Pothen arrived at engineering by way of a small rebellion. His parents preferred medicine. He preferred law. Mechanical engineering looked like a sensible bridge to patent law, a sturdy degree to carry while he planned another life. Then, during his second year at the New Jersey Institute of Technology, he wandered into a friend’s entrepreneurship lecture. A door opened. Pothen became less interested in arguing over inventions and more interested in making them happen.
The timing was severe. It was February 2020. Pothen and a friend wanted to start something, though they had no idea what. A month later, the question changed from “What company should we build?” to “What can we get into people’s hands?” A student team had entered a competition to design protective equipment. Their entry did not win. They kept going anyway.
The CommonHealth Project turned a modest face-shield design into a distributed production network. Clear sheets, foam, elastic and staples went into kits. Volunteers assembled them at home. Drivers handled contactless drop-offs and pickups. Finished shields traveled to institutions across northern New Jersey. Pothen secured materials, packed them and managed distribution in Bergen County. He later described his role as scaling and optimizing a network of 200 constructors, drivers and distribution managers.
There is something clarifying about a supply chain built from kitchen tables and car routes. Every abstract noun becomes a physical consequence. “Capacity” is the number of kits assembled by Thursday. “Distribution” is a driver at the correct address. “Quality” sits inches from somebody’s face. Pothen had found the logistical side of building products, and the lesson was not subtle: a useful idea is merely the beginning of the work.
“I’ve become a big believer in building unconventional paths for yourself.”Mark Pothen, 2022
The engineer who kept changing the assignment
Pothen did not fit the familiar childhood-tinkerer script. He was candid about struggling early at NJIT, including tears of frustration over Calculus II. Yet engineering gave him a way to turn uncertainty into projects. At NJIT’s Mobile Medical Care Unit, he worked on commercialization: refining a possible business model, helping with grant applications, developing partnerships and evaluating research. For his senior project, he designed a scaffold-like electric lift intended to help wheelchair users enter and exit cars.
The projects varied, but the underlying move stayed consistent. Step into an awkward system. Find the real constraints. Build a bridge between the technical thing and the people who must use it. Engineering, in Pothen’s version, was not an identity test. It was permission to investigate.
Axuall became his formal education in software product work. As a student business analyst, he helped standardize a release schedule for software used in workforce credentialing. After graduating in 2022, he joined full time as an associate product manager. More roles followed: product work at FindOurView and OculoMotor Technologies; product and go-to-market assignments at Cape, PayAngel and Steadywell; a product executive-in-residence stint at Techstars.
A résumé can make this sequence look restless. Seen as apprenticeship, it is unusually coherent. Pothen kept studying how young organizations decide what to build, how they get it into a customer’s hands and what happens when a neat product diagram meets the disorder of a working day. The labels changed. The operating question did not.
Six months inside the work
Before co-founding Beacon Health, Pothen spent six months embedded in a primary-care practice, owning parts of its operations and augmenting them with AI. The placement mattered because the administrative work was not one task. It was a chain: identify what needs attention, open the relevant system, gather context, take action, document the result and follow up when the clean path fails.
The operator’s loop
This became the foundation for Beacon, which Pothen started with CTO Obinna Akahara in 2025. The company’s agents operate inside existing electronic record systems and carry workflows from start to finish. Its launch materials emphasize quality-gap work, documentation review, patient outreach, referrals and prior authorization. The product proposition is deliberately unshowy: no new portal for a busy team to remember, no extra dashboard demanding tribute. The software enters the systems already in use.
The practical intelligence here is not simply knowing which button to click. It is knowing why the button matters, what must be true before clicking it and what exception arrives next. Pothen’s six-month immersion supplied a kind of operational grammar. Akahara brought experience building production AI systems. Together, they framed Beacon as a pair of digital hands attached to a process, rather than a clever answer floating above it.
By Beacon’s public launch in early 2026, the company said it was live with an independent physician association supporting 40,000 patients. Y Combinator selected the company for its Winter 2026 batch. Those markers give the story scale, but they do not change its original shape. A group has work that needs doing. The work crosses people and systems. Pothen wants to make the handoffs hold.
Beacon’s choice of language is revealing. The company calls its software “AI employees,” a phrase designed to shift attention from conversation to completion. A chatbot can answer a question and leave the next five steps to a person. Beacon is attempting to own the sequence, including the dull middle where information is moved, a record is updated and somebody discovers that the obvious path has produced another exception. The claim is operational, so the standard of proof is operational too: did the work finish correctly?
That standard suits Pothen’s history. The CommonHealth network could count finished shields and completed deliveries. Axuall had release schedules. A practice has queues that either close or remain open. He has repeatedly chosen environments where progress eventually becomes visible, however complicated the route. AI adds uncertainty to the mechanism, but it does not excuse ambiguity in the outcome. For a founder who learned on physical supply chains, the digital version still ends with an accountable handoff.
Relationships are infrastructure, too
On a startup interview, Pothen summarized one of his operating principles in six words: “Be kind and expect nothing in return.” He argued that relationships compound over time. People encountered in one setting reappear as partners, sources of talent, introductions or customers in another. The return cannot be scheduled, which is precisely why generosity cannot be treated as a transaction with nicer stationery.
That view fits a founder whose first serious building experience depended on volunteers and mentors. Pothen has publicly credited civil engineering instructor Vatsal Shah, NJIT’s makerspace leaders and VentureLink general manager Will Lutz with shaping him. During college, he also spent time around the Amnesty International chapter, the debate club and the student newspaper. His education happened across rooms that did not share a syllabus.
It also helps explain his preference for long-term, positive-sum collaborators. Beacon’s business requires entering organizations with established routines and asking people to trust software with consequential, fussy work. Technical ability gets a meeting. Reliability earns the second month. Relationships carry the rest.
Build before certainty arrives
Pothen’s advice to students was to prioritize building things, even the small project someone else dismisses. The money and job, he said, would follow. Find a mentor. Take risks. Reach out. The prescription sounds almost suspiciously straightforward, though his own path supplies the fine print. Building reveals taste. It identifies which problems remain interesting after the novelty departs. It also produces evidence when a person’s original career story stops fitting.
His original story featured law. The revised one moved through engineering, community logistics, commercialization, product management and startup operations. Beacon is the current chapter, not a grand conclusion. Its ambition is to reduce the administrative machinery that drains time from small practices and to return some ease to the working day. Pothen and Akahara still have to prove that their agents can perform reliably across old systems, varied organizations and the inevitable exceptions.
But the method is visible. Pothen learns by entering the process. He treats manual work as field research and a relationship as something to tend before it is useful. He keeps enough of the engineer’s instinct to map a system, enough of the operator’s patience to respect its edge cases and enough of the former would-be lawyer to make an argument about how things ought to work.
The accidental lecture did not give Pothen an idea. It gave him a better habit: follow the work until it teaches you what to build.