By: Elena Vasquez
Daniel Cohen-Dumani arrived at his first job in Switzerland certain that software was the work. Thirty years, four technology waves, and two companies later, he says the only skills that never expired were the human ones.
The young man who reported for his first day as a junior consultant in Switzerland had a fresh computer science degree and a clear idea of what he had been hired to do. Technology was the job. He would build things, fix things, and be measured on how well they ran.
He was, by his own account, wrong about almost all of it.
“I learned more in those first two years than in the entire degree,” Daniel Cohen-Dumani says now, three decades and two companies later. “And almost none of it was about technology.”
What he learned instead, at the firm that would become Accenture, was a set of skills that had no line in the curriculum. Listening before solving, because the question a client asks is rarely the question that matters. Earning trust before advising, because nobody takes a hard recommendation from someone they do not trust, however right the analysis. Choosing useful over impressive, because the elegant solution nobody adopts is worth precisely nothing. “The relationship isn’t the soft part of consulting,” he says. “It’s the load-bearing wall.”
The lesson would have a long shelf life. Cohen-Dumani went on to found Portal Solutions, a technology consultancy in the Washington, D.C. area, and grew it from a team of one to 60 over fifteen years. Along the way, the tools of his trade turned over completely, by his count, five or six times: mainframes to client-server, then the web, the cloud, mobile, and now artificial intelligence. Every technical skill he acquired eventually expired. The human ones never did, and he credits them with letting him ride each wave instead of being retired by one.
Scaling the company taught him a second lesson he now passes to younger founders: the job changes underneath you, and nobody announces when. At first, success meant the work was excellent, and his name was in every deliverable. Somewhere past ten people, his hands left the product, and success came to mean the team was excellent, a transition he calls the hardest of his career, because everything that made him good at the first job fought him in the second. Past thirty or forty people, even direct leadership, stopped working, and the role became designing the systems that would run without him in the room. “Nobody promotes you to founder twice,” he says. “You relearn it live, usually a year later than you should have.”
He sold the company to a large accounting firm in 2017 and stayed on inside an organization of 2,000. The morning after the sale, he recalls, was harder than the negotiation. Everyone congratulates you, he says, and nobody warns you about waking up without the thing that had your name on it. He describes the feeling not as regret but as unemployment of purpose, and it took him longer than he likes to admit to name it.
The cure arrived in November 2022, when ChatGPT launched, and Cohen-Dumani concluded within days that the industry he had spent his career in was about to be transformed. He chose, as he puts it, the disruption seat over the disrupted one, gave notice a year and a half later, and founded Experio, which builds AI that lets professional services firms find and use decades of their own expertise. The problem it targets is one that bothered him for thirty years: firms that know more than they can find, paying twice for lessons already learned. “This one bothered me so much,” he says, “I eventually built a company and product about it.”
What strikes him most about the second act is how much the founder’s life has changed since 2002, when a week off was theoretical because the business slowed whenever he stopped. This past summer he stepped away for a real stretch, with AI agents handling the routine work in the background, and nothing stopped. “Founders get to have lives now,” he says. “That’s new.”
He has also discovered, somewhat to his surprise, that his most-read public writing is not about wins. A few months ago he began sharing lessons from his career online, and the posts about the messy parts, the morning after the sale, the jobs he was slow to let go of, are the ones strangers write to him about. His generation of executives was trained never to show the seams, he acknowledges. He has decided to keep showing them, on the theory that polish sets a standard but struggle offers company.
Asked what he believes about the coming decade, he returns to the two-year lesson that started everything. Artificial intelligence, he estimates, will eventually do 80 percent of consulting work. The remaining 20 percent is judgment, empathy, and the instinct that reads a room, and it stays human. The tools will keep changing, he says. Bet your career on the parts that don’t.
