As a math professor, how can I help grad students find nonacademic jobs?

19 points by impendia ↗ HN
I am a math professor at a second-tier state university, and many of our Ph.D. students decide (rather sensibly, in my opinion) to avoid the academic job crunch and instead to seek positions in industry.

I would like to help them, but I am at a loss as to how -- I don't have much in the way of useful advice or contacts.

The local job market is not very good, so I'm interested in what I might be able to do to help students with a national search. Any suggestions?

24 comments

[ 3.3 ms ] story [ 46.1 ms ] thread
If you don't have any industry connections, you may not be of much help. But don't feel bad, it's your job to educate your students, not get them jobs.

There are a few places that can help your students find jobs, like the ones listed here: http://www.ams.org/profession/employment-services/employment... But once you look at the listings you'll see that they are mostly academic job listings.

In general, the companies willing to hire pure math people are large companies that have a lot of money to spend on developing new things. So I would give basic advice to your students to apply to large companies. Small companies want a more rounded person and math PhDs don't often come with enough practical skills (unless they are good at programming or did applied math).

The US military also employs a lot of mathematicians and if that fits in with their moral values and they can get a security clearance (read: not foreign and no criminal record) then there is some decent money to be made there. There are a decent number of non-defense government jobs out there as well, but are usually in some applied field like economics/finance or statistics.

But really, as a professor, if you aren't able to hook your students up with a head hunter or personal industry contacts, they may be better off getting job help from someone else. If you have no industry experience yourself, the best option may be to just keep from offering any advice in the hopes of not steering your students in the wrong direction.

Thanks. But could you recommend a "someone else" I might steer them to?

Our undergraduate CS and math majors are not strong (and the local job market even more so), and I suspect that our career services office has adapted accordingly.

I really have no idea, but off the top of my head, perhaps you could invite some companies for talks or visit them to check with them if they could use any mathematicians.

Maybe some could use maths but don't know it yet.

> perhaps you could invite some companies for talks

That might be an interesting idea. Might go nowhere, but could be worth trying.

What area of math are you in?

(After googling): Number theory. So obviously NSA is an option. They hire grad students for the summer, encourage your students to do this if it interests them. Also encourage students with programming chops to spend a summer doing an internship at a big firm that owns their own low-level compute stack, which will necessarily include crypto components. Intel, Apple, Microsoft, GPU manufacturers, ARM, Broadcom, etc. It may be hard to get your first few students in, but once you place a few and get introduced to their bosses you'll have contacts and it'll be easier.

Cast a net outside your field, too. Some students are studying number theory because it's interesting, but aren't wedded to doing it forever. It's worth talking with them about what they're really looking for and nudging them to take some CS / stats / finance courses if they have wider interests.

Finally, as a former math grad student who left to go to industry 10 years ago, thanks for even thinking about this. Most professors don't, and you're already doing your students good.

I'll send you an email with some more suggestions.

OP's profile says analytic number theory
> So obviously NSA is an option.

Indeed, my first Ph.D. student was very much a self-starter, wanted very much to work for the NSA, and landed an internship and then a full-time gig. He asked for rec letters from me, which I was very proud to write, but he basically did everything himself.

>It may be hard to get your first few students in, but once you place a few and get introduced to their bosses you'll have contacts and it'll be easier.

An interesting thought, thanks. That said, I will probably end up advising around 10-20 students over the course of my career (and that's if I'm lucky! Most around here advise fewer, our program is small).

That former student would be an excellent resource with which to put your current students in contact.
I think the best thing you can do is encourage them to take math related courses that can lead to non-academic jobs, such as statistics, machine learning, data science, bioinformatics and quantitative finance.
Get in touch with your previous students, both those in academia, and those in industry.

If they're in academia, ask about their industry connections. If they're in industry, ask if they can help you. If you can't get direct contacts, at least find out what skills and knowledge is being looked for, and what options exist.

It's very easy to forget the power of an 'alumni' network. Who we know is often our greatest asset.

Have them do homework in Haskell. Find local consultancies in the area where they can work part time to get industry experience.
As I said, unfortunately I don't think that there's any local company which I know of and which I'd recommend to a talented student.

This place is very, very much not Silicon Valley.

The best thing is to give your students the freedom and encouragement to do internships during their PhD/MA/MS. That'll help them find outside mentors, experience, and contacts, as well as in-roads to future employment
Force them to take programming classes. Like intro programming -> data structures -> advanced data structures. Focus on making them develop practical programming skills. If they can do that, and have the iq + discipline to do well in math, there will generally be software development jobs available to them.

But seriously, and this is going to sound arrogant, but oh well -- you should consider the ethics of taking 5+ years and an enormous opportunity cost (probably easily $0.5mm) to train students for a job they aren't going to get. Even if students are focused on math, there are areas of math where industry wants phd trained students: applied stats and optimization both spring to mind.

This is the best advice. Tell them to take a couple of programming courses. Give them some simple projects like drawing a Mandelbrot set.

Or more seriously, advise them to do something other than getting a PhD, such as getting a Master's in 2-3 years and taking a couple of computer courses along the way.

> and this is going to sound arrogant, but oh well -- you should consider the ethics of taking 5+ years

Not arrogant at all. But --- these students decided to pursue a Ph.D. when they applied to graduate school, i.e. before they met me. And if they approached me about writing a thesis, then they know full well that their thesis topic won't translate directly to industry. (Applied stats and optimization are outside my expertise.)

I most certainly don't want to be that asshole who tries to intimidate Ph.D. students into not leaving when they're unhappy, or into seeking academic jobs when they'd rather do something else.

But, I don't see it as my job to discourage students from following in my footsteps.

Give them environment to develop projects where they have to learn Math Languages on their own. Require that they choose their language; try to advise matching problems to languages.

R, Stan, Figaro, Venture, Hakaru Coq, Cryptol Sage, Magma, Maple

Give them environment to develop projects where they have to learn Math Languages on their own. Require that they choose their language; try to advise matching problems to languages.

Bugs, R, Stan, Figaro, Venture, Hakaru, Coq, Cryptol, Sage, Magma, Maple

Try sending them to http://versatilephd.com

Even if you can't personally offer any assistance, being willing to discuss a non-academic future lets your advisees know that it's OK to seek alternatives. That was definitely not the case when I was working on my doctorate, and I stuck around for a couple years longer than I would have if the alt-ac track hadn't been considered a failure.

If you don't have many industry connections then help them pick a thesis committee that has more of these kind of connections.
the majority of companies (especially in research) do not recruit without internal reference. additionally, most research companies have a known hiring problem, in that they arent able to fill their talent requirements.

find companies that are interesting for them, then contact them. ask them for their input on this. most'll likely suggest cold applications, but it may grab someones attention

Math PHD students can target Investment banks/ hedge funds. Ask student's to follow Quantnet/ wall street oasis on how to break into wall street as a Quant.
>I don't have much in the way of useful advice or contacts.

Go to the Advanced search feature on Linkedin. Sort for alumni from your "second-tier state university" with current titles of CEO, CTO, CIO, CFO, COO, VP. My hunch is that very quickly you'll have a robust target list of influential decision-makers spread nationwide, even overseas.

Next reach out to these people individually. (*Yes, it's time consuming, it's worth it.) Try to have a live phone/Skype conversation with them. Make a friend; explain your mission to help Ph.Ds find jobs, ask for their advice. BTW, this is how you tap into the hidden jobs market.

Bonus, you will grow your own personal network & broader university community influence in turn. Additional food for thought on this process, read Adam Grant's book Give & Take >http://www.goodreads.com/book/show/16158498-give-and-take