I will finish my PhD in Applied Physics (simulations of quantum transport and light scattering in nanowires) in around 6-8 months and I have already made the decision of leaving Academia as soon as I graduate. The main reason is that I consider that my publications record after the PhD will not be good enough to be able to compete for good positions in the post-doc world. I have also lost the motivation and my interest in the field due to the lack of satisfying and rewarding outcome (I tend to blame myself for this, but I had almost no supervision during the last year). If I stay in Academia I will most likely be jumping from post-doc to post-doc contract in mediocre/low level research groups for the next decade.

I have recently started to consider pursuing a career in Data Science: job offers seem to be abundant and the salaries are good, plus it seems that it is possible to work in many different fields once you are experienced (which is something I love since I'm interested and curious about many, many things).

I currently lack the core-set of technical skills a data scientist should have (R, Python, SQL, Statistics, etc.) and I will not have that much time left to take on-line courses until I write and defend my thesis. However I have strong experience working with other programming languages and I know that I could easily get familiar (not talking about proficiency) with everything needed in short time. After all, everything should be conceptually much easier than the theory of my PhD/MSc and I also have a good background in Statistical Mechanics.

Do you guys have any advices or experiences to share on how to make this transition? I am a bit afraid of the real world after I finish. Also, what would be the most optimal way to get a data scientist job after the PhD? I have thought of looking for an internship, but I do not know if I could just do it by taking on-line courses and doing things on my own.

Thank you very much in advance!

Edit: As it may be relevant, I am not in US but in Europe

  • 5
    An anecdote; A student of mine who is graduating with an MS in applied mathematics that has involved R, Python, and data analysis and who has taken other steps to prepare for a career in data science reports that interviewers have told him that they have hundreds of applications for each data science position, most of them from people with graduate degrees in other disciplines. I fear that the moment may have passed when it was easy for anyone with a PhD to get such a job... Commented Apr 4, 2019 at 18:41
  • 6
    You are competing with undergrads who may have done four years completely focused on data science plus one year of industry experience. There are even a lot of highschool students with those credentials. Is there no way to combine your existing set of skills with data science?
    – Fraïssé
    Commented Apr 4, 2019 at 18:44
  • 4
    Consider looking at machine learning rather than (or in addition to) data science -- perhaps more overlap with your skillset and more R&D type positions. If you are in the US, consider small defense contractors doing R&D (maybe start by looking for companies in your desired area that won SBIR grants on interesting topics -- the winners are publicly posted).
    – cag51
    Commented Apr 4, 2019 at 19:00
  • 4
    @Brian Borchers: As another anecdote, the person I was talking about writing letters of recommendation for in this comment has an MA in math and knows several languages (including R), and spent over a year trying to transition to a data science position (not just applying for jobs, but putting independent data projects he's done online and has taken several courses in an online data science Masters program he enrolled in). He finally got a job in Chicago a few weeks ago. Commented Apr 4, 2019 at 19:46
  • looks like a "future-proof" job :-) Commented Apr 4, 2019 at 23:18

3 Answers 3


There actually is sufficient demand for data scientists and engineers that there are specific 7-8 week long postdoctoral training programs. Such a program could be a good idea, both as a way to build skills and to show that you've actually completed a relevant project. I have no first-hand experience myself, but I know a former physics postdoc who successfully attended the (free) Insight Data Engineering program and now works as a data scientist. (They also have a program focused on data science.) Another such program is The Data Incubator, and I imagine there are more out there.


Here is a theoretical physicist phd graduate who did the transition into data science within few months before graduation. Currently he is Senior Computer Vision Engineer at Level5, Self-Driving Division, Lyft Inc. & Kaggle Grandmaster : Interview with Kaggle Grandmaster, Senior Computer Vision Engineer at Lyft: Dr. Vladimir I. Iglovikov

TL;DR Invest in learning datascience and machine learning by doing kaggle competitions.


It is not difficult to learn Python if you are experienced in similar languages (mostly MATLAB) and a background in statistics is very useful (as you mentioned statistical mechanics). I agree that you can compete in Kaggle, although it may be hard in the beginning. For sure you need to acquire hands-on experience, it is one of the most important requirements in industry. You can start from easy data science projects in Python and proceed fast. Build your own portfolio which is what you are going to outline a lot when applying for jobs and during interviews.

You must log in to answer this question.

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