I am a few months short of completing my Physics PhD. My thesis work has mostly been on mathematical/physical modeling of biological systems, combined with simulations (i.e., no use of ML). However, I have completed several personal projects over the years (on GitHub and my CV), where I have tried to demonstrate my knowledge of both models/architectures, and programming and software engineering skills. I've had friends in industry look over my CV, code review my projects, and I feel prepared for technical interviews. My PhD institution is considered prestigious, and my GPA is high, if that matters.

Over the past 2.5 months, I have applied to 150+ jobs (mostly Data Scientist, Quantitative Researcher/Analyst, ML Research Scientist, ML Software Engineer roles), but have not landed any interviews. I know that not having industry experience is a huge weakness, especially given the current job market. My goal is to find a job where I can expect to occasionally work with tools more complex than regression. Given that I can't do anything about the job market, I see a few next-steps as options, and was hoping to get some feedback on which ones would be more worthwhile to invest in in order place me in a better position to eventually get there.

  1. Internships. The easy way to get experience, but so far I have not had much success with these applications either. Is being overqualified a concern? Is competition for internships also currently very high?
  2. Start-up. I have an offer to join a startup founded by some friends. The work would be more data analysis, however, with probably no more ML than regression. I'm worried that experience in this role may not count for much (and the pay would be precarious).
  3. Post-doc. Aim for a post-doc where I have the opportunity to publish in ML conferences. This might make me a better candidate for ML Research jobs, but would it be enough to overcome the lack of industry experience?
  4. Keep applying. Maybe I am not a bad candidate for an entry level job and just need more numbers on my side?

In general, I'd really appreciate any information on how a hiring manager would perceive me, now and after one of these stints.

  • 2
    I have to imagine this question will be closed as opinion-based. However, I would like to say: 2.5 months is not a very long time to be job-hunting in todays market. Keep your head up and don't get discouraged about that no matter what path you end up taking.
    – InBedded16
    Commented Jul 12, 2023 at 17:06
  • I've studied physics (did lattice quantum field theory simulations & high performance computing) and now work as a research scientist with a deep learning company. I've written a blog post about my particular transition, perhaps that is helpful to you. Commented Jul 14, 2023 at 20:00

3 Answers 3


I suggest you just keep applying for the jobs you want, and if money is tight then take on some other role you're not worried about leaving if you land a job where you want to be.

Sooner or later you will get a chance to showcase your skills. Also network as much as you can in the industries you're interested in, it's much easier to impress one person who respects you than a faceless hiring manager skimming CV's. Most of my higher level work has come looking for me through my network.


You might consider looking for a specialist recruitment agency that deals specifically with vacancies that suit your skillset or perhaps ask around people working in your academic field for some advice based on their own experiences.

There are many institutions that hire PhDs without industrial experience - physics or mathematics doctorates into finance is one example.

In the meantime if you love your work then do indeed try a postdoc if you can find one as it will help you stay up to date in your field.

It really is just a matter of time.


I started out with a PhD in biomedical engineering (focussing on computational simulations), then after that fell apart I went into government data science. In that latter career I've been on several recruitment panels, reviewed a couple of hundred written applications, and interviewed several dozen candidates. This advice comes from that perspective and some things will be weighted differently in industry.

If I were assessing you as a candidate, based on an application similar to your question here, my notes might look something like this:

  • PhD, math/quantitative focus
  • Light on ML experience but quantitative skills would probably translate - expect they could pick it up. [Government is perhaps a bit more willing to take an attitude of "find somebody with aptitude and train them", which might not be an option for a startup that's in a hurry.]
  • Hard to evaluate soft skills - unclear whether has experience working in teams, communicating outside workplace situation, managing timelines etc.

In that context, the most likely reason for my passing you over for interview would not be your lack of ML experience but the lack of info about your soft skills. It's not necessarily a deal-breaker at grad level, many candidates at that stage just haven't had a lot of opportunity to demonstrate these soft skills.

But if I only have capacity to interview N candidates, all else being equal, the ones who can show me something in that area will make the shortlist ahead of the ones who can't.

Again, this is something that government probably weights higher than industry (as a generalisation - neither of those sectors is homogenous!) but being able to demonstrate those other skills is unlikely to hurt in any sector.

When you do get rejections, it's worth asking for feedback; many places don't give it, but it can be useful when they do.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .