28

I would like to ask for help since I am on the verge of finishing my PhD Math degree but I am worried that there are almost no jobs which I qualify for. I am an international student in the US and my field of research is in cryptography but my programming skills are not up to par of course to CS majors (I only know how to program on a basic level in Python, C++). Basically, my skill set is more of a pure math major and not of an applied math major.

My worry is that since there are almost no academic job opportunities out there for me since I see that most of the academic positions available are in applied mathematics not named cryptography and if there are cryptography jobs out there, certainly I am not the most qualified since I see myself as not really a computer science major.

When I look up for industry jobs available, it's either exclusive to US citizens (like In the NSA) or those jobs are looking for applied mathematics majors like those who know data science, statistics, machine learning among others.

I am already "old", I'm turning 33, having a mediocre resume that does not stand out and as time goes by, it seemed that taking the PhD route is a bad decision.

Any advice?

12
  • 29
    I don't understand why you immediately compare yourself to CS majors. More theoretical maths is valuable in a number of industries, starting with insurance, banking and logistics and certainly not ending there. Where are former students of your advisor working?
    – Roland
    Jan 26 at 5:53
  • 9
    I suggest you ask your advisor about his other graduates.
    – Roland
    Jan 26 at 6:49
  • 6
    Even if your advisor didn't have other students while you were there, presumably there were other pure math students in your department that needed to get jobs, yes? (In general, yes a lot of academic positions are aimed at "applied" math nowadays, but pure math graduates still find jobs.)
    – Kimball
    Jan 26 at 13:23
  • 5
    Is retaining the right to legally live in the US - and hence getting a job that can and will sponsor a visa - one of your concerns? If so, (1) I apologize for our unjust immigration laws, and (2) please mention it in your question because it is relevant. Jan 26 at 16:22
  • 7
    You could explore blockchain engineering and the general cryptocurrency space? You have a massive advantage over other developers by being able to understand whats going on at the mathematical level, moreover the industry is red hot. I personally know a couple of a people in this space that have made an enormous amount of money (via salary not even speculation/investment). This would leverag your cryptography specialization heavily and make your PhD "worth it" more so than if you had done any other subject. But doing the crypto PhD shouldn't preclude other paths Jan 26 at 20:16

11 Answers 11

26

it seemed that taking the PhD route is a bad decision.

Maybe it was, but you can't change the past. You can only change the future.

Things to do now:

  • Figure out what else you learned. It might seem like you didn't learn anything, but you must have, since earning a PhD is not trivial. If you didn't use Python or C++ for your PhD, what did you use? If you studied cryptography, do you also know how to secure a system against hackers? If yes, you could try searching for jobs in IT security.
  • Approach your university's career center, preferably sooner rather than later. They'll be able to offer more personalized help.
  • Look up jobs that require a degree in your broader field. In your case, that would be mathematics.
  • Consider searching for non-US jobs. There is no requirement you stay in the US.
1
  • 33
    Note that not all university career centres are competent. For a physicist I know, when asking her French university career centre what one could do with physics, all they advised was teaching physics. Simultaneously, the Dutch tax office came to my Dutch university to recruit physics students before they had even graduated (surviving the first year was considered sufficient to qualify).
    – gerrit
    Jan 26 at 12:57
23

Crypthography is a hot topic and security experts are very scarce. You might not quality for the actual job offerings, but you might find some employer willing to invest some months into you, to get a loyal and qualified worker.

There are many jobs to design systems, understand what signatures and hashes are for, avoiding bad patterns. That is what an average programmer is not capable to do. This will be your job!
These problems occur in everywhere: cars, planes, satellites, weapons, coffee machines, communications.

1
  • 10
    "security experts are very scarce", indeed, though I think "cryptography experts" are even scarcer, I know of organisations who need people who understand cryptography to validate the security solutions being touted by numerous companies. Jan 27 at 17:16
15

Doing a PhD in mathematics compared to programming at a normal level is like comparing flying a jet fighter to driving a car. You might not know how to drive the car, sorry program, but you would have no problem learning it if you wish to.

As very few math PhD-s are expected to apply for jobs, it is very seldom written as a requirement or even as a positive thing. But send out your CV to a bunch of the places you might like to work at -- the worst that can happen is that they say thank you. Searching jobs is like marketing a new flavor of ice cream, if no customer knows it exists, no-one is going to ask for it. Same for you, no potential employer know you exist until you tell them. So, as the saying goes, just do it.

4
  • 5
    I know several math Ph.D.s whom I would be loath to let anywhere near my code. And my Ph.D. is almost a math Ph.D... (probabilistic combinatorics / complexity theory).
    – einpoklum
    Jan 26 at 19:20
  • 3
    @einpoklum Agree. There are a lot of people I would not let near my code either. But in a production environment you follow the rules and most people can learn that if they get a bit of training. It might not end up as "masterpieces" but then professional programmers are not supposed to make that.
    – ghellquist
    Jan 26 at 21:12
  • 11
    Most people can learn to be mediocre-to-bad coders which still churn out something usable. But few people, as they're taught that, learn to be good coders. And writing good code is often contradictory to the immediate instincts or interests of commercial environments...
    – einpoklum
    Jan 26 at 21:38
  • 1
    I have around half a dozen stories of students fairly successful academically until they had to code, and this number is growing. Seeing how it actually broke people, more than once, makes me strongly disagree with "but you would have no problem learning it if you wish to", sorry. Other than that, good advice. It's all about marketing your skills and finding solutions to someone's problems (which they sometimes didn't know they had).
    – Lodinn
    Jan 27 at 23:24
9

Allure's answer has some great advice, I won't repeat it.


Have you considered any analyst roles? In particular in government.

I was probably in a similar position after completing a Neuroscience PhD and after taking up programming and a bit of web development, applied to be a Civil Service Analyst (UK). Here, the civil service don't look at your career history, the interview is more like a test.

But, from all the analysts I've met, many of them come from PhD's, mostly due to a broad understanding of statistics and bits of programming. It might not be glamorous, but it pays the bills and problem solving is probably more important than your technical skills.

Python could lead you towards Data Science rather than straight up statistics.

7

Let me offer a satirical spin on other answers, comments and opinions of the majority:

  • You can try to become a teacher at a second-rate college in the middle of nowhere. However, with over 500 applications per spot, your chances are very slim (link).
  • You can try to get into data science, but with plenty of competitors who have more relevant qualifications, you may need to complete another degree before you can even start thinking about it (link).
  • Of course, there may still be some companies left willing to hire mathematics graduates for standard coding roles. However, given your lack of experience, most professional programmers will not let you anywhere near their code, so you might be out of luck (link).
  • Of course, you could always try to become a cashier or a gas station attendant, but with the looming new wave of the pandemic and ever-increasing store automation, the odds and time will be against you.
  • But fear not, there is no requirement for you to stay in the US. As they say, the (3rd) world is your oyster (link).

P.S. Oh, and don't forget to check out the incompetent staff at your university career center, who will surely have no clue about how to help you to find a job (link).


I sincerely hope you do not take any of this too close to heart. The truth is that with a research-level degree related to cryptography you should be in a much better position on the job market than almost anyone else (with the exception of ML experts), provided that you are willing to diversify slightly and pick up some new skills along the way (link).

3
  • 1
    try to become a cashier or a gas station attendant -- You didn't give a link for this one, but probably one can be found by using the search word "overqualified". FYI, I found myself "on the street" so to speak in January 1988 (failed Ph.D. exams 2nd time where was at) and needing money for food and rent, and was unable to get any such job, largely I think due to having an M.A. in math (and not currently a student) and thus being considered overqualified (despite previously having many such jobs, 3 in fact at fast food restaurants in the past 7 years where I did quite well). (continued) Jan 27 at 17:57
  • Maybe things are easier now, at least in the U.S. due to the huge shortage of employees for such jobs, but if one goes up the ladder a bit, I suspect there are still obstacles to being "overqualified". However, in the case of the OP, a specialization in cryptography seems to me about as ideal as one can get (at least in math), so I suspect the OP's outcome might eventually be similar to this person's experience (see Edit at end). Jan 27 at 18:06
  • @DaveLRenfro Yes, the world is a funny place. I have known too many people with stable careers in mid-level management roles in banking/technology with the most mediocre qualifications and seemingly the most mediocre technical abilities. Yet someone like Yitang Zhang had to live in a car and work in a Subway shop after completing a Ph.D. in mathematics :). Their life worked out eventually. But I wonder how often it simply doesn't.
    – Anon
    Jan 28 at 12:19
4

Your plea for help sounds like a classic example of needing to adjust your perspective and expectations. What were your expectations when you started your PhD? What were your goals? If you just wanted a good job and future in the USA, then take courage because the world is still your oyster. You just need to change your perspective.

Have you tried looking for jobs in "less prestigious" universities? If you look outside BIG city universities, there are plenty of middle of the road universities that maybe are in less famous locations but are hidden gems and you may have to teach some but you'll have a cool low stress job that is yours for the keeping. Not to mention, these locations are usually lower cost of living and you end up financially set for life at a relatively young age.

5
  • 8
    Nope - all permanent academic jobs are hard to get nowadays. I would guesstimate Grinnell College got somewhere between 500 and 1000 applicants for their job. Admittedly, that might be considered prestigious by some measures. Jan 26 at 18:41
  • @AlexanderWoo In this job market, landing a permanent position as a young academic might not be the greatest option indeed, but that doesn't mean "temporary" positions should not be considered. Now, in the US it might be different but the fabled tenure track is not the only good career opportunity at or around academia unless one goes the purist route.
    – Lodinn
    Jan 27 at 23:20
  • @Lodinn - Permanent jobs, tenure track or not, are getting lots of applications. Temporary full-time positions also get a fair number of applicants, and there is a marked bias in hiring towards people who got their PhDs somewhat recently. Ten years past PhD, you have to either get a permanent job, which has 500 applicants, or you can try to live on part time jobs that pay very poorly, or, as most people in this situation choose, you can leave academia and start over in a new career field. Jan 28 at 0:09
  • @AlexanderWoo Fair enough, one would need to settle down ten years after the PhD, but with these ten years of work experience, would one also not be better equipped to do so? Similarly, leaving academia does not necessitate a fresh start - some people do side gigs in the industry R&D while still in academia. A sanity check of "will my skills still be marketable enough to feed myself and (if applicable) my family in ten years" is much needed as early as possible, of course.
    – Lodinn
    Jan 28 at 0:25
  • I'm admittedly biased - wages in academia here are barely livable so working in academia AND in the industry or freelancing at the same time is rather the norm... And the industry doesn't seem to care about the PhD any much; it is only concerned with you delivering results. Now THAT is a marketable skill.
    – Lodinn
    Jan 28 at 0:30
2

A math PhD plus some basic experience in programming is enough to get a good job as a data scientist. Python is the key language for machine learning which is another plus. You might not qualify for a job at Google but that is because everyone wants to work there. There are a lot more companies out there looking for people doing data science and data analysis that there are people qualified to do it.

Your math PhD is generally seen as sufficient proof that you are one of the people qualified for that or at least can easily be training to be. Apply to lots of data science jobs, point out your math PhD and don't worry if you don't satisfy all the stated experience requirements. This is a wish list and in this field most companies can't get everything they want.

2
  • 9
    I'll disagree with this answer. As degree and certificate programs in data science have sprouted up in recent years, employers have begun to demand some formal training in data science. Jan 26 at 18:43
  • 2
    Agreed with @BrianBorchers this isn't such an easy path to get on now. Back in circa 2016-2018 it was doable Jan 26 at 20:20
1

You may find that there are research organizations interested in Cryptography at both the really large software companies (Microsoft, Google, Facebook, ...) and at some of the "Security" companies (RSA, etc.). I expect the competition for these positions to be fierce though.

Try to find some names at these companies. In many cases, they will be publishing papers just like they would do if they worked at a university. Read their work. Try finding reasons to ask them questions about their work; building up a network is likely the best path forward.

1

As someone who has a Ph.D. in EE but does research in cryptography, I am going to say that there's plenty of space to do research cryptography as long as you can bring a nuance of what you know to the field. Over on https://crypto.stackexchange.com, I primarily answer questions regarding semiconductor implementations as that's the nuance I bring to the field. Also, it's worth mentioning that I do not do classified work. This is an important aspect because I often work with international students as collaborators, whereas, many of my colleagues who do not draw that line often cannot discuss their work or travel.

Back to the original question: I believe that there is a space for anyone who can contribute to any field, but you'll have to show that you can add some value. See if there's something in https://eprint.iacr.org/ that overlaps with your interest. That will give you leads on companies and groups who could possibly use work and where you could contribute.

0

Tons of things to do, crypto [e.g. zero knowledge chains] and finance stand out. Even privacy for internet marketing etc. Obviously also all the other tech firms if you want to brush up your programming and data skills.

1
  • I have some notion of zero knowledge proofs. Is that related to zero knowledge chains? What is a zero knowledge chain?
    – Galen
    Jan 28 at 16:59
0

In addition to what has be written, two ideas:

you can try out some internship positions. There are many for data-science and probably not a few in engineering or cryptography. Less tech-oriented, the big investment banks offer internships as well.

If you decide on going into data-science, spend a few months learning practical and theoretical stuff; in an interview, your mathematical understanding will score you points above the average applicant. For example, I doubt the average data-scientist today could easily read the book Elements of Statistical Learning - try it out.

Second, consulting firms take PhDs quite often into consulting/analyst roles (though they take MBAs in a higher ratio). If this is foreign to you, you can try out some practice case-interviews to get a feel, you might like it.

As is often the case, you'll benefit from having a connection to put in a good word. I'd suggest workshopping your CV and polishing it for every position.

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