A few weeks ago, I was laid off from my position as Director of Controls and Electrical Engineering. The separation was abrupt, coming at a time that I was preparing for a relocation from New York to California.

Losing my job in such a way was firstly, a psychological shock. I questioned my self-worth, my professional acumen, etc. While, it is always good to be introspective, it is also good to know your own value, which for me, after over 20 years in my field, (and despite some lingering impostor syndrome) I can say was objectively well-represented by the positions I held. That being said, there are always going to be risks of things beyond our control, and that’s just a normal part of life I’m working on getting better at accepting.

Beyond the psychological shock, there were other immediate, practical questions about income, insurance, and the next opportunity. Focusing on these immediate questions was important for me personally, because I was able to use it to empower myself, rather than sinking into rumination spirals. Other, harder questions I had were: How should I describe what happened? What kind of position do I actually want next? How do more than twenty years of experience in materials science, particle accelerators, instrumentation, control systems, electrical engineering, operations, and technical leadership fit into the opportunities that employers are offering today? With a background as niche and eclectic as mine, these were not easy to answer.

That being said, my response was/has been to approach the job search as an engineering problem.

That did not mean I needed to treat people, careers, or hiring decisions as deterministic systems. A job search is nuanced. It contains noisy data, incomplete information, delayed feedback, and a great deal of human judgment. But I can still use the same basic principles I use when approaching complex technical systems (“when you’re a hammer, everything looks like a nail…”). The approach I took looked something like:

AI as a Cognitive Engineering Tool

Artificial intelligence, among other modern applications, has become an important tool for me in the job search process. I do not use AI as a substitute for technical knowledge, professional judgment, or personal authorship. But I do use it as a cognitive engineering tool: a way to organize large amounts of information, identify patterns, challenge assumptions, and convert loosely connected thoughts into structured material that I can evaluate and revise.

For me, that distinction was/is important.

As we’re seeing more and more, an AI system can produce polished language while still being technically wrong, strategically misguided, or inconsistent with the person using it. Consequently, every output requires review. Therefore, I always verify technical claims, and often have to remove exaggerations, correct context, and rewrite language that does not sound like me.

The value for me is not that AI “does the thinking.” It’s that it helps expose my thinking to inspection.

During this job search, I have used AI to:

In that sense, AI has functioned less like an automatic résumé writer for me, and more like a combination of requirements-management system, research assistant, technical editor, and structured notebook. I actually used a few AI tools like Claude, Chat GPT and Gemini, and referred to them as my “Executive Team”. It was a fun experience, giving them executive roles with me as the CEO.

Translating Experience Across Industries

One of the most difficult parts of a senior-level job search was not determining whether I have the relevant experience. It was determining whether an employer will recognize that experience in their own, often unique terminology.

For example, my background is centered on EPICS-based scientific control systems, accelerator operations, precision instrumentation, distributed computing, machine protection, diagnostics, and high-availability infrastructure. A data-center company may instead describe its needs using terms such as BMS, EPMS, telemetry pipelines, historians, OT/IT integration, OPC UA, BACnet, or fleet-wide controls management.

These are not identical systems, but many of the underlying engineering concerns are closely related:

AI has helped me build these conceptual mappings. I can give it a job description, compare the terminology with my own project history, and ask where the overlap is genuine and where a gap remains.

That last part is essential. A useful analysis must identify both: help me determine if there’s enough overlap to be viable, or whether there’s an actual gap. Therefore, I often have it help me decide what the “fitness” level of the position is based on percentage.

For a role involving embedded systems, for instance, I can point to experience with real-time control, PLCs, FPGAs, watchdogs, heartbeats, event counters, configuration scrubbing, interlocks, IOC recovery, hardware/software integration, and failure-mode analysis. At the same time, I should not claim years of hands-on development in a particular RTOS or microcontroller family if that is not what I did. That is where the fitness test plays the important role. For most jobs I apply to, I look for a fitness of 80% or better.

So while AI helps me articulate the boundary, I remain responsible for keeping it honest.

Building a Feedback Loop

The job search has also become an iterative feedback system.

Each application generates information, turning it into a data science problem. Rejections may indicate a weak match, an internal candidate, excessive competition, location constraints, compensation misalignment, or simply bad timing. Often times, that root cause is unknown, without any feedback beyond the rejection itself. An interview, when I’ve been lucky enough to get one, provides richer feedback: which parts of my background attracts attention, which explanations resonate, and where interviewers ask for greater depth.

The process resembles a control loop:

  1. Input: A job description and available company information
  2. State estimation: My experience, constraints, interests, and current pipeline
  3. Decision: Apply, defer, or decline
  4. Execution: Tailored résumé, cover letter, application, and outreach
  5. Measurement: Response, rejection, screening, interview, or offer
  6. Adjustment: Refine positioning, search terms, materials, and preparation

The feedback is imperfect and often delayed, but it is still actionable. As a side note, my average statistic is about 10% of the jobs I’ve applied for, with 80% or greater fitness, have actually gone onto the interview stage.

Since the layoff, I have applied across scientific facilities, national laboratories, semiconductor technology, energy, aerospace, embedded systems, industrial automation, AI infrastructure, and data-center engineering. Several applications have advanced to interviews, including panel and second-round discussions. One nearby opportunity progressed rapidly through an initial conversation and a multi-person technical interview process. Other opportunities have involved fusion energy, national laboratories, accelerator facilities, and large-scale controls and telemetry organizations.

Not every application has moved forward. That is expected. The objective is not to eliminate rejection; it is to build a sufficiently strong, well-targeted pipeline that no single decision controls the outcome.

The Human Work Remains Human

AI can help organize a career, but it cannot decide what makes the career meaningful.

It cannot replace the experience of operating a scientific facility during a difficult failure, coaching an engineer through an unfamiliar problem, balancing schedule pressure against safety, or deciding when a technically elegant design is too fragile for production.

It also cannot fully understand the personal stakes of a job decision.

A position may offer excellent compensation but require relocation. Another may be close to home and allow me to remain near my children and established support network. A role may look impressive on paper but move me away from the kind of technical systems I enjoy. Another may contain a few terminology gaps while aligning deeply with how I think and lead.

Those decisions require my own human judgment, values, and responsibility.

AI helps me collect the relevant facts and examine the tradeoffs. It does not, and cannot, choose for me.

Learning Through Application

One unexpected benefit of this process has been how much I have learned.

When a posting references a platform or concept outside my primary experience, I use AI to explore it interactively. I ask for the underlying architecture, failure modes, terminology, and comparisons with systems I already know. I then verify the explanation against established, relevant documentation and determine whether the connection is legitimate.

This has helped me deepen my understanding of embedded firmware, RTOS scheduling, power electronics, building-management systems, electrical power-monitoring systems, industrial telemetry, data-center cooling, nuclear instrumentation, and other adjacent fields. My previous background involved much of these systems and metrics, but with different applications to other industries.

That knowledge has been a more powerful use of AI than simply generating application language. It turned the application process into a technical learning program for me.

The goal for me was not to memorize enough vocabulary to survive an interview. It was to understand unfamiliar systems well enough to have an honest engineering conversation about them, and see how well I would fit into a role where I was required to work with and/or manage them.

What I Have Learned So Far

The first, and probably most important lesson I’ve learned, is that a layoff does not erase the work that came before it.

The systems still operated. The projects were still delivered. The teams were still led. The difficult problems were still solved. The end of one employment relationship does not invalidate the knowledge and capability I accumulated over decades.

The second lesson is that senior engineers, myself included, often have broader experience than their job titles communicate. Translating that experience requires careful consideration, an open mind, evidence, and an understanding of how other industries describe similar problems.

The third lesson for me was that AI is most valuable when paired with skepticism. It accelerates synthesis and processing, but it must be reviewed like any other engineering artifact. Good use requires verification, configuration control, clear requirements, and an accountable human owner.

Finally, I have learned that the job search itself can be productive work. It has literally become my full-time job, complete with my “AI Executive Team”. The search has forced me to inventory my experience, revisit projects I had almost forgotten, clarify the kind of leader I want to be, and explore technical domains that connect naturally with my background. It has also made me re-evaluate my own self worth, and do some personal soul-searching. I have genuinely considered switching career paths during this period, but have decided that continue pursuing opportunities where I can be the most effective, as that level of productivity and effectiveness is what satisfies me most.

I would not have chosen the disruption that began this process. But engineering has never been about controlling every input. It is about understanding the system in front of you, responding intelligently to changing conditions, and building the next state with knowledge and experience.

That is what I am doing now.

← Back to Articles