I have two engineering degrees — a BSME in 1975, an MSNE in 1984 — and I worked as a nuclear engineer for less than a year. I never worked as a mechanical engineer at all. What those degrees actually got me was credibility in a field I never formally trained for: software. For some reason, "nuclear engineer" on a résumé opens doors in IT that a computer science degree doesn't. I've never fully understood why.
The work itself started at home and mostly stayed there. My first job, in 1975, was remote before anyone called it that. From 1991 to 1996 I ran a database consulting firm out of Kirkland, Washington, building custom systems for clients I mostly never met in person — Apple, the Electric Power Research Institute, Genentech, Puget Power, Synopsys, the University of Washington. From 1997 to 2008 I wandered as an IT project manager through a string of large organizations — EMC, the Methodist Hospital System, Texas Children's Hospital, the U.S. Army, Los Alamos National Laboratory, General Motors, Hitachi Data Systems, Cisco — never staying long enough to put down roots, learning something different at each one. Then sixteen years managing iOS and Android app development for clients around the world. Since 2008 I've lived on a beach in Ilhéus, Brazil, running what has, over time, become a company of one — now one person and a set of AI collaborators that didn't exist five years ago.
The common thread across fifty years isn't the technology. It's the people. I've worked with colleagues and clients on every continent, almost entirely without meeting most of them face to face, for longer than "remote work" has been a phrase anyone needed. It reminds me of engineering school in the early 1970s — the place was already full of people from everywhere, a mix of nationalities I didn't think much about at the time and have thought about a great deal since. I've spent my whole career recreating that hallway, one client and one continent at a time.
This year I went back to the actual physics. My master's thesis, written with my advisor and published with him in Fusion Technology in 1986, worked out how neutrons leak through the gaps and seams in reactor shielding — the places where two slabs of steel don't quite meet. I reproduced that same 1984 calculation this year using OpenMC, open-source Monte Carlo software that didn't exist when I was a student, and resolved a convergence limit the original 1986 peer review had flagged as unsolved. The manuscript is still working its way through journal review as I write this.
It's an odd thing, sitting back down with a forty-year-old problem and finding the tools have finally caught up to the question. But it's not really a departure from the rest of the career. The degree was never really about doing nuclear engineering for a living. It was about learning to think in a way that turned out to be useful everywhere else — including, apparently, in a version of engineering nobody had invented yet when I was in school.