From Academia To Industry and (Halfway) Back

This is the story of how I went from academia to industry and (halfway) back. I touch the topics of job satisfaction, introspection, why I decided to do an industrial PhD, what that means, how it compares to traditional PhDs in academia, and some tips for people considering doing one.

Outbound Flight: From Academia to Industry

It’s Summer 2019 and there I am in a Toronto basement, frantically writing a research paper and my Master’s thesis, while also preparing for job interviews. The warm, long summer days were spent mostly in the damp comfort of the basement.

Two years of research work boiled down to a couple of months. On top of trying to be a competent researcher, I’m also trying to show hiring managers I can be a competent employee. Despite the stressful situation, I’m at peace with the life decision that put me in that situation: in Fall 2018 I made the choice to not continue my path in academia and try my luck with an industry position.

Long story short, I somehow powered through and was able to publish, graduate and get a job. That made the last few weeks of my summer much more enjoyable.

I remember in conversations with colleagues I would often say that I didn’t discard coming back to a PhD in the future. I don’t think they really believed my words. But I don’t blame them, I don’t think I believed my own words either. Anyways, life was good and I packed my bags to start a new chapter.

Layover: A Short Stay in Industry

I landed a position as a software engineer working on a domain I cared about. That’s as good as it gets for a first job. I was not doing research (I tried), but I was still solving (somewhat) interesting and impactful problems.

At around the 6-months mark I started to feel unfulfilled by my job. It wasn’t a concrete thought on my mind, but the feeling was definitely there. Don’t get me wrong: the job was challenging and it made me grow professionally in unforeseen ways. But ultimately I was lacking a sense of accomplishment. Simply put, I wasn’t solving the problems where I felt I could contribute the most. Those problems lived solely in my head, and occasionally saw the light of day over a coffee break with an interested colleague. It was from this realization that I now extract a core lesson from my experience: being passionate about the business domain does not guarantee job satisfaction. There’s many dimensions contributing to job satisfaction: relationship with managers and colleagues, work environment, team dynamics, salary, etc; and you shouldn’t rely on any particular one to justify your decision to stay (or not) at a particular job. For about a year I rationalized my feelings:

“Maybe the job I’m doing is not fulfilling, but I can still say I’m working on this cool project, and at some point I’m bound to work on something I like more, right?”.

That sounded reasonable to me, until…

My team was moved to an entirely different project. I took it harshly, but it wasn’t the end of the world: we were in the middle of a pandemic and I still had my job (as opposed to many people who lost theirs). It happened to me and it can happen to you, so you just pick yourself up and consider your options. For me it became a major issue since I started losing the motivation to excel at my job, and that would severely hinder my performance in the long run. In any case, I suggest doing the thought exercise of listing the different dimensions of job satisfaction that affect you the most, and know where you stand. Know what things could change (new manager or team lead, change of projects, …) and understand how it can affect you. I am not saying to disregard change altogether; oftentimes you will be positively surprised. Instead, I would say you should “try to make it work” but also be mindful of how this change fits into the roadmap (see below). If the change is incompatible with your goals, maybe it is time for you to grab the proverbial bull by the horns (your career) and turn it in the direction you want.

And so I did. Time to re-build my career.

Inbound Flight: Meet Me Halfway

My job search was done from a most privileged position, I already had a job. Not everyone has that luxury and all I can say for those in a tough situation right now is: hang in there, work on yourself, know your value.

After some introspection, it became clear to me what my next steps would be. I wanted to do research. I wanted to solve new problems. I wanted to solve real-world problems for which we don’t have great solutions.

Here are some tips and guiding questions in case you find yourself in a similar situation as me back then:

• Know yourself: Where do you want to go next? Where do you want to be in 3 years? Where do you want to be in 5 years?
• Draw a roadmap: What’s needed for you to reach your 5-year goal? What path puts you closer to your goal? Are you lacking experience, knowledge, connections?
• Think concretely on the next step: Now that you have a roadmap, what’s a reasonable next step? How does it fit into the bigger picture?

I want to emphasize something on the last point. Life is unpredictable and not every decision will bring us closer to our goals. What’s important is being cognizant of the goal and not lose sight of it, even in troubled times.

I went through a similar introspection exercise as the one I outlined above and decided that my next step was either to go get a PhD (train myself as a researcher), or find an industry position that allowed me to do more open-ended research or at least work on applied research. With renewed hope for the future, I started searching and preparing for potential interviews.

Fast-forward 3 months, and I found myself in a crossroads: 1) do a PhD at a university or 2) do what’s informally known as an industrial PhD, a PhD within a company.

I will briefly mention their main differences (disclaimer: these facts come mostly from my personal observations and research about robotics/AI programs in Europe, so it may not apply in all cases)

• Employer: in academia you are an employee of the university. On the other hand, industrial PhDs are hired by a company. The company has a partnership with a university and there’s a contract binding the three parties: student, company and university.
• Hiring process: there’s a lot of variance here. Typically, university applications compromise sending documents (statement of purpose, research proposal) and then interviewing with faculty. Industrial PhDs may go through a hiring process very similar to full-time employees (HR screenings, research presentation, technical assessments, etc). From my experience, academic hiring in AI/robotics is starting to resemble industrial hiring (potentially due to high volume of applications).
• Collaborators: in academia you will be typically interacting with faculty, undergrads and other grad students. In industry you will interact mostly with employees of the company: scientists, engineers, managers, but also with the appointed faculty advisor. Cross-collaboration between academia and industry is possible in either case.
• Responsibilities: in addition to research, in academia you may also need to write grants, take courses and teach. In industry you may need to integrate research into products (see below), write patents and other various company-related activities.

Those are the core differences. Further differences highly depend on the nature of the industrial PhD. So far I’ve identified two poles describing a spectrum:

• Product-based: the PhD student will conduct research with the main goal of enhancing or developing a product or service for the company. Academic dissemination is a byproduct of this work, and in collaboration with the academic partner (faculty from university)
• Research-based: the PhD student conducts research with the goal of creating new knowledge and publishing results in papers and journals. Product development and patenting is a byproduct of this process and is done in collaboration with employees of the company.

Again, this is truly a spectrum of the type of experience you may expect from an industrial PhD. Now I will outline some guiding points about doing an industrial PhD (it largely applies to the PhDs in university as well).

• Know the expectations: understand what’s expected of you in terms of publishing and product development. Does it align with your roadmap?
• Know your collaborators: understand what team you are joining, who are the members, how it fits in the company and with whom you will collaborate. Know your team’s strengths and how to leverage them.

Back at the crossroads: I did my roadmap, I did some introspection. I was now equipped with the knowledge to best inform my decision.

I went with the industrial PhD (it’s research-based). So you could say I’m halfway back to academia, doing research in a company with academic collaborators. The decisive points for me were: 1) interesting research topics while staying close to industry problems, and 2) work alongside experts in the field.

Closing thoughts

There’s no winner in the academic vs industrial PhD debate, and it is silly to think there is. Some programs will be a better fit than others, it all depends on your personal motivations. That’s all there is to it. Do your research. Know yourself. Be confident about your decision and enjoy the ride.