Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: Who hires mathematicians?
165 points by gremble on Dec 6, 2017 | hide | past | favorite | 142 comments
Granted, as I am only graduating with a straight BSc it is perhaps a bit presumptuous of me to consider myself a mathematician, but I would like to. I've been looking at jobs and the only people who seem interested in me are banks or people looking for a "quantitative analyst" in the financial sector.

Who hires mathematicians, other than the aforementioned financial industry?

I know that machine learning is pretty math heavy, and I have taken a look at some of the mathematics involved and some programming firms also don't mind if you have a BSc Mathematics/Applied Mathematics degree. But doing that doesn't seem like doing mathematics.

This is perhaps an odd question for the site, but I have been struggling with this and everyone here seems professional and helpful from my years of reading here.

Not me, that's for sure. I interviewed two maths phd's years ago, who wanted to get out of academia because they felt themselves above chasing grant money, but then didn't want to do anything that wasn't their phd topic or related (of which I didn't even understand what it was, and they couldn't explain, or give examples of what it could be used for).

On the other hand, I did hire people who studied maths in programmer-ish roles; meaning: they didn't have to be programmers, we'd teach them, but they did have to apply maths to our concrete problems and translate it into programmed solutions. I don't think any of them used particularly advanced maths, or did anything they could get published in maths journals (not that I'd recognize it if I'd see it)

So, from my perspective, to be employed as a 'mathematician', you have to go into academia. Otherwise, you have to apply your math skills in some other field; but you won't be called (or probably feel) a 'mathematician'.

(all people with advanced maths degrees I know work as programmers)

That was my experience. All my math degree ever did was mildly impress potential recruiters, but I had to learn how to code things unrelated to mathematics to get a job.

The only job I was able to get as a mathematician "outside" of academia was a brief stint teaching high school mathematics. Which was fun and I would do again if the pay were higher.

It was fun? Could you talk about that some more?

Any time someone talks wistfully about a job they'd like to have, it's good to listen.

I was replacing someone on mat leave so I could only have the job for one year. I was teaching three regular classes of geometry, one honours geometry, and one IB Standard Level class which was a whirlwind mish-mash of topics.

The honours class was an absolute blast to teach. The kids were engaged, they were able to handle the toughest problems I could throw at them, they liked being there. Having the privilege to teach that class was worth the whole year.

The regular geometry classes were more challenging but still lots of fun to teach. It was more difficult there to get across to the kids. Many of them didn't want to be there. I still tried and came up with lots of different ways to explain things. Not all of my methods were successful.

The IB SL class was a mixed bag. These were dedicated students who worked hard but didn't really want to take a lot of mathematics, because they didn't want to take the Higher Level class. Some of them understood, some of them didn't. I was very sad to have one student who for whatever reason I just couldn't reach. She seemed dedicated, very polite, she wanted to succeed, but she failed nearly every exam. It was a little heartbreaking. I have no idea what I was doing wrong.

I was very young and green. I have had more informal teaching experience, mostly in the way of giving tech talks or explaining computer things to other computer people, but I always wish I could be explaining geometry again to kids, trying to nudge them to explore and learn in their own ways.

This is quite anecdotal and not very helpful to the question asked.

All the time people apply for jobs with different expectations than what the job role is. That is what the interview process is for, so both sides can find out if there is a good fit.

"That is what the interview process is for, so both sides can find out if there is a good fit."

Well yeah, that's exactly what the anecdote illustrates - people who aren't hung up on 'doing maths' or 'being a mathematician' will have ample opportunity; those that are better like it at universities because that's where they're heading.

I graduated with a BSc in Mathematics in 2008, and didn't exactly know what I wanted to do. I just typed the keyword "analyst" into job search engines. It worked pretty well! I ended up getting a job in a field that I didn't even know existed when I started (web analytics). It turns out that almost every industry has some sort of data that they can benefit from having someone analyze it, and a Math degree is a pretty good qualification for it.

Don't worry too much if you don't match the exact description of a job posting. I've actually never met all of the job requirements for any of the jobs I've been hired to. Point of fact, most people will look at a Math degree as a degree in "Very Smart" and give you the benefit of the doubt about being able to pick up any specific skills you might be missing. This is doubly-true for niche fields (like web analytics) where there aren't specialized degrees, and picking up the skills on the fly is a requirement.

I will say, almost any job that you take won't look like "doing mathematics" in the sense that you're used to from college. That pretty much doesn't exist outside of academia, with the possible exception of research labs that require an advanced degree.

I wouldn't say people with BSc in math are "very smart," they are in my opinion wickedly, awesomely super-smart. I have taken some math courses which contain material that math BSc:s go through during the first semester and those were much harder, by a large margin, than any other courses I have taken!

But I can't imagine going through three years of that and then have to spend your time with web analytics. It must feel awfully pedestrian. The level of the math can't be anywhere close to what a BSc is capable of.

Meh, most of everything is pedestrian.

I like that idea. Trying a slightly different approach, and searched on indeed.com for "mathematics" and "$50,000" (filters for jobs that are supposedly 50K USD and up per year).

Jobs on the first page:

- Mathematician @ Reflexion Investments

- Entry Level Systems Engineer @ Boeing

- Special Agent @ FBI

- Quality Engineer 1 @ Northrop Grumman

- Software Quality Engineer 1 @ Northrop Grumman

- Junior Independent Credit Review Officer @ Bofi Federal Bank

- Mathematics & Statistics Student Trainee @ US Air Force

- Academic Tutor @ MathWizJohn Tutoring

- Cost Analyst, Junior @ Booz Allen Hamilton

- Police Recruit (Entry Level) @ City of Chula Vista, CA

...the list goes on.

Snap! I'm also a mathematician who works in web analytics but I graduated high school is 2008! I agree with everything you have said but I will add that you SHOULD KNOW HOW TO CODE and NEED TO KNOW EXCEL AND SQL (this is A LOT EASIER than knowing how to code)

Another Math grad here. Currently working in video game analytics. Lots of options for us it seems :)

I'm a mechanical engineer with strong minors in applied computational mathematics and scientific computing, and math has been, for me, a super-power. It opens all kinds of thermal, structural analyst jobs. I can optimize all kinds of stuff. I can automate simulations and plot the results in ways that the non-technical can get a good idea of the right direction to go. That all comes from math.

Can you hand-code a solver for Maxwells equations or Navier-Stokes (lid-driven cavity flow)? That opens electronics and thermo-fluids. How are you with classic and singular perturbation methods? That gives you signal integrity at Intel - they pay pretty well. Get a few languages under your belt - icky things that are gold plated. MatLab, LabVIEW, Python, SQL, and C. That gives you most of mechanical engineering, lab-based data collection, tons of current "data science", being able to work with previous content, and making your code go really fast, respectively.

Try not looking for jobs on monster, or dice or such. Get friendly with a technical recruiter at Manpower Technical and ask them to get you a few decent contract positions to help you both return some excellent value, and to grow your professional breadth. Make sure some of the positions are business, production, design, and leadership in that order. If you do leadership first without the others, you are wasting yourself.

Read a few books on how to negotiate salary. Your earnings at age 25 determine your total lifetime earnings, so if you let yourself get low-balled early, it can cost you a few million in total lifetime earning. You don't want that.

Once of my heroes is Karl Kempf. Read what he, applied mathematician that he is, has done. He returns a defensible $8 billion per year every year in new value to his company. He is someone to emulate.

> Can you hand-code a solver for Maxwells equations or Navier-Stokes (lid-driven cavity flow)?

For my undergrad, I mostly focussed on number theory and differential geometry. These topics are mostly useless unless you want to work at the NSA, which only works for the US. Most of my undergraduate curriculum is absolutely useless except for the art of mathematics itself.

For my Masters I wanted to learn more about numerical analysis, hoping to land a job this way. I suppose that did help me understand topics like machine learning, which really seems to me like numerical analysis rebranded, but I didn't end working on anything like it. In fact, what happened is that I started contributing to Octave and used the experience from working on that to justify getting other programming, non-mathematical jobs.

The reason I'm telling this story is just to emphasise that most mathematics is useless, always has been, always will be, just like most art is useless. You mentioned very specific areas and applications of mathematics, but mathematician in general will always have difficult finding jobs pertaining to their mathematical interests and need to diversify towards the need of the job market.

I think the comments here could do with being put in a little perspective. There are programmers reading HN (well, all programmers) for whom one important way they can improve as a software engineer is to improve their powers of abstract thinking. You people who have studied advanced mathematics, any of it, are exceptionally skilled at thinking abstractly. I think there’s a risk of underestimating how useful that is in software engineering.

I think it's a bit superstitious how we think that being good at math means you'll be good at programming. I'm not complaining, because that superstition helped me. Like I said, my math degree did give recruiters an initial good impression of me. Other than that, I don't think it translates into much of a benefit when writing code. I have personally witnessed many mathematicians write the most atrocious code, myself included.

Also, if it really were about that kind of structured thinking, maybe just teaching chess to kids would be just as well.

Wouldn’t you say there’s a difference between structured thinking (say, a kind of breadth-first search of ideas with pruning) and abstract thinking (which allows a programmer to, say, identify the need for a Form class on the backend to model and process POST data)? Bad programmers will just keep writing similar code to process the POST data with no abstraction about what is being done and why. Related, mathematicians use lemmas and theorems to reuse thought.

I think the relationship is as tenuous as being a good cook or a good chess player. Coming up with a unique recipe for cake or a good strategy for checkmate is about equally comparable to writing good code.

OK, I don't think I agree. In the area I work in, writing good code isn't so much about coming up with a clever sequence of moves; a large part of it is being adept at modeling real life situations with abstractions in software, and being adept at manipulating and combining those abstractions, and seeing when and to what extent they can be used more broadly, and knowing how to assess whether the end product is correct to the required degree.

Yes! Remember, FORTRAN survives because mathematicians still use it...

I definitely feel the same way about my double-major in Math. My favorite super power that came from the degree is probably "being able to read a statistics text book". I turns out that lots of industries run on statistics. And, if you can read a boring text book that covers the statistics used in that industry, you suddenly understand why people are doing what they're doing in that industry; making you a domain expert.

Anything you can link to to read more about Karl Kempf? A quick google hits linked-in and paywalls.

> Your earnings at age 25 determine your total lifetime earnings

Do you have more info on this?

I'm interested in this too. As a 27 year old, if this holds true, it's very, very frightening.

I have a PhD in mathematics, and have been working in the public sector (NASA) for a little under 10 years. A mix of interest in numerical linear algebra and software engineering helped me work into a niche in scientific computing and engineering software development.

Math is a very, very broad professional interest area. It branches into education, academia, and several areas in several industries around the senior year of an undergraduate program. I'd recommend undergraduate math majors try and get into co-op or paid internship programs as soon as they feel comfortable with it. But that's advice that I'd also give to ANY student these days.

Most technical organizations will hire folks with a mathematics background or entertain the though for a promising applicant. But there are almost no job titles out there called "Mathematician". That can be confusing to some.

Any chance you know someone hiring for said scientific computing roles? I studied Applied Math / CS undergrad, have been gearing up to work in Autonomous Vehicles, but really any challenging numerical work would be interesting to me.

I have a PhD in maths (applied, though most of my undergraduate was pure), and lectured for a couple of years before moving to a bigger city and deciding to break out of the academic hamster wheel.

My wife already had a job secured, so I had the opportunity to be a little patient. This really paid off.

I considered many of the types of jobs mentioned by others.

Analyst / tech-consultant jobs didn't seem technically interesting enough. Mostly seemed to involved adapting the same framework or analyses to companies' slightly different requirements.

I got through to an on-site interview with an algorithmic trading firm offering huge bucks, but decided I didn't like the idea of moving money around to make very rich people even richer. My dad still says I should try and do something like this for just a few years then change to something more fulfilling.

I was interested in the betting / odds playing type companies and even went to interview with one (rejected). In hindsight I don't think I'd have enjoyed it as much as my current job.

After a month of my CV being online, I was rung personally by the MD of a smallish company that makes software for public / municipal services. He described excitedly the types of optimisation problems he was hoping to solve, and said he'd been looking to hire the right person to work on it for years. The personal touch and his obvious enthusiasm and energy sold it, and I've now worked there for a year. I've learnt and implemented a lot of algorithms and techniques from a broad range of topics - network routing, discrete optimisation, graph theory, markov chains, plus some bonus stuff like cluster analysis and signal processing. It's been more fascinating than all but maybe the first year of my PhD.

I guess my advice is just that these types of jobs are out there. I noticed a couple of other comments about radar tracking and remote sensing, both of which sound awesome. If you have the opportunity, hold off until you find something that strikes you as exciting or even unique. If you don't have the opportunity, perhaps find something "rote" but keep an eye out.

That's interesting, my PhD is in an engineering field and projects have included a lot of what you mentioned ('graph theory, markov chains, ... cluster analysis and signal processing'...). I just went through a job search and couldn't find anyone very interested in these skills, so I'm back in another postdoc. Any advice on companies/industries specific to those types of methods?

There are a lot of interesting problems in vehicle routing, route optimisation and GPS data processing. There often aren't good libraries which solve them with enough flexibility and performance, and better algorithms are being researched and published all the time. Thus it's quite common for companies to develop in house solutions. A blog post by food delivery comoany Ocado [0] was posted on HN a while back, and I'm sure companies like CityMapper and Strava that work with maps (as well as Google and Bing of course) are constantly trying to improve their algorithms.

[0] https://ocadotechnology.com/blog/ocado-internet-of-vans/

I will keep this in mind when researching jobs.

wow, graz!

Srsly that sounds really nice and very similar what i would love to do.

From my experience, I would offer the generalization that nobody hires mathematicians. Mathematics in society is more of a skill set than a professional title. The most challenging or cutting edge math that could be commonly used is the LINEST() function in Excel. A person who is good at math also has a lot of great skills to offer a company, it is just selling those features and not the calculus.

I started out as a math major, then I transitioned to a double major math AND stats because stats is more applicable. I struggled for a year looking for work (also US immigration sucks, even for Canadians) and ended up in a master degree program in Industrial Engineering. I chose engineering specifically for the word "engineering". I was lucky that I discovered the field of Industrial Engineering at that university otherwise I was headed for a BS in Mechanical.

Continuing formal math education will further limit the kind of jobs you can apply, increasing the level of competition. Even the BS in Math left me with the feeling people saw me as over qualified, lacking regular skills.

Math is super great by the way, just not the idea of being a "mathematician". It (unfairly) causes alienation of your true potential.

This is my sentiment as well. I have BS in math and MS in applied math and now work as a data analyst. Problem solving and abstracting problems to general properties and attributes are the most worthwhile skills that my math education has provided me, at least as marketable skills. I learned R and python on my own and no doubt this has been a huge help in finding jobs, and has allowed me to be closer to the tech at my jobs versus a more traditional excel based biz analyst role. This has allowed me to learn a lot more on the job, like being able to directly touch dbs, create my own ETLs, learn AWS products, etc.

You (OP) may take for granted the way that your math education shaped your brain to think about things. Don't. This mode of thinking is one of your biggest assets that employers are after.

What is the difference between mech and industrial?

Mechanical engineers make things, industrial engineers make things better. The discipline started in the 1900s mostly concerned with improving manufacturing efficiency and cost. Now it has grown to improve all facets of product quality and business process (physical and digital). It is one of the best disciplines to use applied math. https://en.wikipedia.org/wiki/Industrial_engineering

Industrial engineers work on production processes. You can't make a bridge better without fully "grokking" the physics behind the original design. It isn't exactly amenable to incremental improvement.

https://www.youtube.com/watch?v=g8UePdbDmMw This was one of the keynote speakers at the 2015 IISE National Convention. Nancy Currie was/is works for NASA and has education in Industrial Engineering. She worked extensively investigating the Columbia space shuttle incident. Her involvement shows how industrial engineers participated in making space flight safer / better. It's really not grokking the physics that can make things better. Excellent presentation.

NSA is the obvious answer if you want to do math - they do math at a lot of levels, from research in algorithmic number theory to programming in mathematically correct exploits. One former NSA employee told me that they retrained him as a programmer.

Data Science is possible, but generally seems hard to break in without prior programming experience or a masters or higher - I had problems getting in as a 4 year PhD dropout from a prestigious math program, and ended up teaching myself programming & have carved out a nice career as a software engineer. My math & physics background has proven to be a bonus in my favor when interviewing, and I generally am favorable to people with a math background I encounter in industry/interviewing candidates since I have found it fairly uncommon.

I would go to on campus career events if possible, and talk with company recruiters about interviewing & general tips. There are potential other options depending on how open you are to them - the options are a lot more open IMO than with most degrees. One thing that might help is to go to Indeed, LinkedIn, etc., and just search for jobs in a particular area - if a particular profession sounds like something that might be feasible/palatable for a career or first step, jot down the title, and continue. This doesn't mean you're committing to anything in particular, but it will help you understand what you are looking for better and better prepare for any future interview sessions where they ask you what got you interested in <insert position>.

I am a mathematics professor (at a non-elite university).

If anyone has advice on how I can help our students (at all levels -- undergrad and grad) get jobs in any of these industries, I would be grateful to listen!

I have a Bs in applied mathematics and I'm a software engineer. I think that mathematicians make incredible software engineers (not me necessarily ) if they do two things, take some basic cs classes, get co-ops/internships programming.

I faced a _lot_ of bias when getting into the industry because people didn't want to train me to program or thought I'd be too weak of a software engineer, etc... and I believe having internships/co-ops would have massively helped rectify that and also have given me a network of jobs to work at.

I'd actually be interested in your experience. Personally I have a degree in Physics. Only two classes in programming (majorly self taught) and work in simulations. How does the software engineering field see people like me? I've applied to programming jobs in the past and just never knew if I was under qualified, didn't stand out, or just slipped through the cracks in the large amounts of applications places get.

And to OP's question, look into simulation. There's a lot I learned on the job, but having basic programming abilities will go a long way. A lot of engineers are weak at programming compared to what I see in CS.

I'm not sure what exactly simulations means? lots of Matlab/R/some specific DSL?

So when I applied to jobs with my applied math degree and 2 years of grad computing experience - but no industry experience - as a software engineer - I got lots of sorry no experience, we can't train you, etc... etc...

I eventually got a job as a support engineer and have kicked around a lot of roles including data scientist and software engineer.

I'd say if you don't have actual programming experience they'd probably see you pretty similarly to me. That being said if you're applying to a giant corporation and don't have some sort of standout crazy resume OR have an internal referral I think you're basically lost in the noise. Applying to smaller startups may get you a second look.

However, now there are things like triplebyte where if you pass their interview you can get an onsite to a lot of companies regardless of your background - which is pretty neat.

> I'm not sure what exactly simulations means? lots of Matlab/R/some specific DSL?

Mostly C++ and python. Though I know Matlab too.

I'm not looking for a job, per say, but it is always good to know what I can get. But you would suggest a head hunter? I do have experience now.

I would suggest finding people you went to school / are friends with at tech companies and ask them to refer you - That is typically a short circuit to getting past the piles of resumes.

I did a double major, math and physics. I was basically self-taught with respect to computers (one CS class on my transcript). I got hired as a programmer by a lady with a Master's in math, who was largely self-taught with respect to computers. That first hire is all it takes - after that, I had experience.

I graduated in 1984, though, so this may not be very applicable to the current situation...

Still the same situation, even for CS majors from less known schools. Somebody has to take a chance, few people want to take a risk on you.

Well that sounds reasonable to me. First job was hard to get though. I can completely understand where the OP is coming from.

The first job was the grace of God. If I interviewed with someone else, I'm not sure that I would have been hired. ("You taught yourself? Sure you did, kid. Run along and play, now.")

Haha that's what I got a lot. Though I think that's why it might be easier to get into programming at an engineering job.

I think that’s true maths provides the problem solving skills

To start - help to get rid of the sentiment that programming is beneath them.

A lot of math undergrads won't take CS classes, won't learn to program, bc they think it is somehow inferior.

source: BS/MS in applied math and didnt learn to program until grad school. I know many other math undergrads that had similar views. I now work in c++ all day - writing algorithms for catheter navigation and geometric mapping of the heart (and previously image processing for factory automation, machine learning/data science for predictive maintenance via aircraft sensor data).

It's a real problem. You don't even need a full CS minor to be useful in industry. An intro programming class, data structures, 1 or 2 algorithms classes. An OS class if you want to be more thorough. In some students it's an attitude problem, in others it's just unfortunate lack of awareness.

There's a grant from the NIH (assuming you're in the US) called the K25. It's designed to get math/engineering people into health related fields. It's for the postdoc level (it might also be applicable for someone with a BS/BA, not sure). It can pay for 3-5 years of salary for the individual to work with a mentor and to learn the field - no knowledge of the medical-related subject is expected since the applicant is explicitly there to learn the field. As such it's widely applicable to math people who want to do research in medical fields.

Caveats: I think you need to be a US citizen or permanent resident and you have to find a mentor willing to take you. This shouldn't be impossible though, since your salary will be paid by the grant so it's not a financial burden and many fields need people on the math/statistics side of things. Look for biostats researchers and contact them even if they're not hiring since again the grant pays the salary they might take someone even if they aren't looking for someone. The application is open three or four times a year, next one in February. I think the October round better matches the typical graduate' time frame, which means students should plan this at least a year before graduating.

As someone going through this process right now, thanks for thinking of your students' careers! I wish more math departments encouraged options outside of pure math and finance. There's definite demand but a difficulty finding the appropriate positions for us.

I’d tell them to begin work in a portfolio of whatever they’re interested in doing. If they want to work in IT then take the time to learn some cloud platform and Linux sys administration. Try to obtain the A+ asap. If they want to be in data analytics then learn python, excel (for the HR departments of the world), and statistics. Then make a code example repo and/or blog with some analytics. Be prepared to start at a very low position and work up as quickly as they want.

That would also apply to anyone wanting to get into those fields. I’d argue a solid math background puts them in a strong place to specialize.

I’ve also always wondered to what extent math would be useful in law school.

"I’ve also always wondered to what extent math would be useful in law school."

I have a software engineering degree (not CS) first, and then put myself through law school part time while working as a programmer, so I have some insight in this. In short, it's completely useless; even programming skills are useless. There are some highly specific (patent) positions where having a technical background combined with law 'pays off'; although if you look at the opportunity cost and the premium you're paid, it's not all that great (and writing patents is incredibly tedious; for litigation, tech skills count for bupkis).

Even for thinks where you'd think having a technical background would matter (IP lawyer, computer crime/forensics legal consultant, cybercrime DA), your tech knowledge is useless. The vast majority of the work there is legal, and the little bit of tech needed is easily explained by a consultant with no knowledge of the legal aspects.

I know (of) quite a few tech/law combo grads (it's actually not all that uncommon), and very few of them manage to synergistically apply both in a career. Those that do, do so because of a third (communication/self-marketing) skill; more than having any technology background.

So, in short: law degree is fun, but don't count on it being anything but a hobby (when combined with tech career).

(oh, I also know a mathematician who's now teaching at a law faculty; he does research on computational ethics and philosophy, that sort of thing. Intellectually very interesting, but 0 real world applicability - in other words, bound to academia - for better or for worse).

I wouldn't think a math background would directly apply to law. I'm thinking it might translate to someone who can build a solid argument. I'm aware that logic in the manner familiar to a mathematician would not apply. It's all about rhetoric and knowledge of the law. That's been my take away from conversations with legal professors/students.

I appreciated your insight. My thoughts were formed with much less knowledge.

In a word: internships. Large corporations and government agencies (at least in the U.S.) do much of their hiring from intern (or co-op) programs. You might want to reach out to your graduates to find out where they work and what they do (alumni offices are pretty good at tracking them down).

The short answer is “everyone”. I’m a mathematician who’s worked in various roles in various sectors. Not one of them actually involved doing any maths, but plenty of them appreciated the kind of mental training studying maths gives you. (There are other routes to this clarity.)

So I guess the question is: are you looking for a job where you get to do maths, or are you just worrying about employability?

I guess not employability as much as hoping someone will pay me to solve problems with mathematics with ideally some interesting problems thrown in. I majored in pure and applied mathematics and I am fully aware the no one is going to ask me to do functional analysis for cash, but it would still be nice to do some sort of applied mathematics for money.

amusingly functional analysis is really important to convex analysis and optimization so yes if you have a phd/masters (maybe) you can definitely get - probably a giant company - to pay you to do functional analysis.

Of all the great replies I've gotten, this one has arguably alleviated my concerns the most.

I've seen at least one or two companies working on optimization at least here on HN. I'd take a guess that Gurobi is one option, and I saw SigOpt post a listing on the recent "Who's Hiring?" thread. Google seems to also do a lot of optimization related work as well. I'm in the same boat, though I was stronger in CS during my studies. The background is useful, but you will only ever use bits and pieces of it at opportune moments.

Data science, data science, data science. The company I'm involved has some blog posts you might find useful:

- https://www.dataquest.io/blog/how-to-get-a-data-science-job/

Feel free to email me if you want to chat more about data science careers, etc (srini at dataquest.io)

I concur. The OP says data science doesn't sound like feel like doing math. On the other hand, job postings for "data scientists" where I live don't sound like they have much to do with programming. They have long lists of math and other theoretical requirements, and throw in "oh, and by the way it would be nice if you had basic Python skills".

If you find out, let me know! One thing I'll caution you about is there's sometimes a bit of a disconnect between "doing math" as perceived by someone who studied e.g. CS versus someone who studied math. For some people, doing math means maybe doing some trigonometry or some basic stats; whereas others won't be satisfied unless they're working on algebraic k-theory or something similarly next level. For people in the former category, there are certainly jobs available with just a bachelors degree and ideally some programming skills, whereas for the latter, you'll almost certainly want a PhD (and even then you may not get to use that knowledge outside of academia, depending on your area of study.) You'll want to find out where you are on that spectrum, and how you feel about grad school / work life balance / etc.

Personally, I got a bachelors in math and ended up working as a software developer. There's enough overlap in the sort of thinking required that makes it reasonably enjoyable. I do wish I had more opportunities to use math in my job though!

> disconnect between "doing math" as perceived by someone who studied e.g. CS versus someone who studied math

In what sense? There are folks studying theory in CS departments who are by all intents and purposes 'pure' mathematicians.

If you see mathematics as a highly transferrable and broad intellectual skill (rather than a narrow profession), everybody hires mathematicians.

In software and computer science especially, mathematicians are likely to find a compelling place to apply and develop those skills.

I've often thought the skillset for mathematics and software engineer are isomorphic. Interestingly, there's a mathematical theorem showing that to be true for functional programming (Google Curry-Howard correspondence).

My observation and experience is that good mathematicians make great software developers. I believe Ryan Dahl dropped out of a PhD program for algebraic topology and made node.js as one of his first open source contributions.

A math degree could be very useful in the remote sensing industry (this has both commercial and military uses). Processing radar and lidar data, construction of 3d representations from imagery and video sources. This has applications in agriculture, mining, medical, intelligence gathering, military intelligence, etc.

In my experiences with companies doing this work it was a healthy mix of CS, math, and physics personnel. The ones working for DoD largely preferred people with PhDs and masters degrees as they bill the government more for those people's time, but they would also provide a lot of financial aid (often 100%) for you to pursue graduate degrees for the same reason.

"Processing radar and lidar data, construction of 3d representations"

these are all 4th year/masters level electrical (or mechatronic) engineering topics at my university (except Lidar although that might come under photonics, which again is an EE subject).

And the way they are taught doesn't emphasize math (well... the hardcore math is mentioned in lectures but not assessed), but rather MATLAB or domain specific software.

Alright, maybe I didn't give the best examples. How about this, office next door to mine is taking radar systems fielded in the 80s, developed in the 70s, and using them to construct 3d models of the detected objects. This is far beyond the capabilities originally intended for these systems and is actually quite impressive (no published results that I'm aware of, unfortunately). This is algorithms and math heavy work. They have a combination of experts (well, some becoming experts) in computer graphics, radar systems, machine learning, and other fields present doing this work.

If you want something that is mostly "doing mathematics", i.e. thinking about Galois groups, differential geometry, or the Riemann hypothesis, that is pretty tough. Many jobs have interesting mathematical content, but to get them you will probably need to convince the prospective employer that you can handle the non-mathematical content, which likely means writing non-trivial computer programs. The one exception that I can think of might be entry-level actuarial positions.

At this point, a lot of people decide to get a masters degree in operations research, statistics, or computer science. But don't completely give up and get an MBA. :-)

I did the same thing, and ended up working for an early-stage biotech company doing some combination of hardware design, signal processing, machine learning, and experiment design. For better or worse, people generally seem to think a math degree implies "smart" and tend to be willing to overlook a lack of any specific skills...

Which company is that if I may ask?

I'd rather not, sorry!

That's OK. I'm looking for something similar, hence the question :)

The NSA. (Really. Some of the largest employers of mathematicians.)

Yes, exactly. Be wary of staying too long - then you may not have much to (publicly) show for years of work.

Not something you'd need to worry about, since you won't get fired from there.

Making yourself illiquid is also a way to cut short your career options. Should you have a shift in life priorities you may find yourself wishing to work somewhere that you cannot. Everyone should be able to say something about their work.

My company which is the largest sports betting company in the Balkans is having a team made of only mathematicians. They do work related to statistics and calculations for ods and game mechanics.

Off topic: one of my secret regrets in life is that I am not smart enough to be a mathematician. I watch the popular channels, I read about the new discoveries, I drunkenly explain some of the fancy concepts to my friends after the fourth beer. I wish I could understand higher mathematics.

"What one fool can do, another can." (Ancient Simian Proverb) - From Calculus Made Easy by Silvanus P. Thompson via https://news.ycombinator.com/item?id=14161876

I have refreshing my mathematics knowledge via videos made by Grant Sanderson.

Search "3blue1brown" on YouTube. I started with his Linear Algebra series (recommended here on HN)and then the Calculus series and finally the Neural Networks series. (highly recommended)

He is working on a Probability Series and I am supporting him on Patreon.

I am going to share a secret with you, I am not all that clever either. I just substitute it with stubborn doggedness to figure out a problem, a willingness to be wrong often and a slight dash of creativity. This has so far made me reasonably successful. If you can get past the necessary, but confusing symbolisms and formalisms, its pretty good.

You are and you can! Mathematicians often do a poor job motivating what they're doing, which creates most of the confusion. There are great examples and intuitive concepts behind most of higher math. Many mid-level mathematicians are either unaware or reluctant to admit to what a great extent their intuition and heuristic understanding informs the more rigorous treatment. There is more to it than the examples, but the examples lead you naturally to the more general case.

Disclaimer: I am a graduate student in EECS, and not math. In particular, I am not a professional mathematician.

At some level, none of us is smart enough to understand higher mathematics (in its entirety). Some illustrations (for which references are easy to obtain via a web search):

1. John von Neumann, in his time, said that he understood 28% of mathematics (no idea where he got that specific number from). Given the 20th century explosion of mathematics, largely driven by the professionalization of mathematics and the drive from the nation state, I can't think of anyone today who can truthfully answer > 5%.

2. One of the favorite, semi-secret pastimes of mathematicians is ranking other mathematicians, e.g is so and so a "first rate, second tier mathematician" or a "second rate, first tier mathematician", etc. The funny thing is that this ranking does not end: Fields medallists fall below the "inner circle" of Fields medallists (who e.g won the Abel prize as well), but even they look up to people universally recognized as singular, but who are no longer alive - popular favorites being Euler and Gauss.

A far more modest task is understanding some aspect of mathematics that one finds immensely fascinating. Fascination and enthusiasm are by far the most critical pre-requisites; without them one can't go far. Often what makes a field of mathematics difficult (at least in popular perception), say algebraic geometry for a stereotype, is the lack of familiarity with the language of it. Language takes time to seep into our brains. Some people can do this far faster than others, and hence appear as "geniuses" when they absorb things without much observable effort.

But that does not make it impossible for others; it usually just means more effort over a longer duration.

There is also a pattern to how mathematics operates. Enough exposure to a variety of topics will demonstrate that many famous, deep theorems, are at their core built out of (in retrospect!) simple, natural ideas - ideas so natural that one can come up with everyday analogs of them. In fact, in a highly simplified sense, most proofs at their heart involve what many would call "high school" manipulations (e.g a lot of arguments in real analysis involve what many call a "3 \epsilon" trick). Yes, there are subtleties involved, and a lot of effort sometimes needs to expended to complete arguments, but the same is true of various fields, including programming/software engineering (just look at the evolution of Unix from the original < 10k lines to today's *-nix).

If one wants something concrete that can be appreciated by many on this forum, consider:

1. http://www.math.ucla.edu/~pak/lectures/Math-Videos/comb-vide... - Igor Pak's excellent archive of combinatorics videos; many of which are for a pretty general audience.

2. Federico Ardila's courses, see above for links; also see https://www.quantamagazine.org/mathematician-federico-ardila... If one checks the course websites of his offerings, there are tiny blurbs with some info about the students of his courses. Many come from quite non-traditional backgrounds, and a non-negligible fraction of them end up doing interesting research.

3. "The Princeton Companion to Mathematics" and/or its sister "The Princeton Companion to Applied Mathematics" - these books do a fantastic job of giving a bird's eye view of the essential unity of mathematics, and can be excellent starting points for diving into a topic suited to one's interests.

Lastly, mathematicians themselves do not always understand what they are doing/have done - otherwise generalizations/refinements that take many years to come up with would have happened much faster. This is especially the case when they (in retrospect!) are breaking seriously new ground, and are thus probing far into the dark. Gauss's earliest proof of quadratic reciprocity used induction on the primes; a technique, although useful, is far from most "modern" treatments; and is arguably not the best way to "understand" it. In von Neumann's words: "Young man, in mathematics you don't understand things. You just get used to them."

Machine learning is actually quite math-lite in comparison to "math" like PhD math. Most masters/PhD math stuff just isn't required or used in the discipline at all. You can get away with undergraduate analysis for pretty much all of it. But it builds off of undergraduate math. So in that sense, you're not really looking for a position for "mathematicans", rather a position for "data scientist" or "quantitative ...", where if you take a field and stick quantitative in front there's a subfield for it. If you search those terms you'll likely find things more in line with what you're looking for.

As for places to look, there's lots of stuff going on in the web and general computing sectors which are now making use of machine learning tools and hiring teams of data scientists. There's also quantitative biology (pharmacology), climate science, etc. disciplines, but many of them want applicants who have a PhD.

If things don't seem "mathy" enough, it's because a lot of the true math jobs and research requires a graduate degree in a math-related field. Doesn't need to be a Math/Applied Math PhD, but even CS, Physics, Climatology, Systems Biology, etc. programs set you up for a math-based career. Without trying to be demeaning, math is a very vertical discipline and the issue is that undergraduate math is really just the basic competences and most of the interesting stuff comes after, which is why many things require a lot more than a BS.

In my experience (undergraduate mathematical statistics, currently "quant financial engineer") this is exactly right. If you want to continue "doing" math in the sense you did in school, or in the sense we assume a "mathematician" does, it really is almost exclusively a consideration accompanying having a doctorate in the field. Mainly because as stated above, undergraduate mathematics really just scratches the surface of core competencies. That being said, while not many fields offer pure math tracks for undergrads it certainly doesn't make it irrelevant. In most "quant" fields mathematical literacy and the thought patterns accompanying having studied mathematics are far more valuable than actual proficiency. Outside of research few individuals are "paving new roads" mathematically, they are most often applying the results of cutting edge research or long standing industry standards in mathematics within the context of their job. This requires a high level of mathematical fluency and comfort with abstract formalisms, not theory expertise akin to that of a phd.

This was very informative, thanks. I am perusing postgraduate studies. I've been accepted into an Honours degree in Applied Mathematics so I am not hanging up my academic hat yet.

My reason for asking the question is two-fold: some badly needed reassurance that I have not rolled myself into a hole (I think it was a design flaw that we cannot re-roll characters in real life) and getting a broader perspective on potential places to keep in mind - in terms of extra-curricular skills that I need to gather along the way.

If it is reasonable for you to continue with postgraduate studies, I would recommend that you do that. The more math you can do or know the better. And it is going to be harder to gain those skills once you leave university.

You haven't mentioned any software skills. I would recommend develop whatever skills you have in that area. The ability to program is becoming as important as literacy and numeracy. Even more so for a mathematician.

My bias is that I have a PhD in math and have been a software engineer for the last 10+ years. I have worked in a lot of different areas and that has largely been because of the combination of math and software (and luck).

Oh. I can program. I've done two university courses using C++. I've taught myself haskell using Haskell Book and for my numerical analysis course I reimplemented all the matlab that we had to do in Julia and Python. A different poster mentioned C/C++ so I am currently considering going through the numerical analysis again and doing it in C++.

I didn't mention programming because I don't see myself being a developer, which I knew would get a lot of attention. I like solving problems with programming, but I don't see myself writing a webapp to collect information for someone to steal.

Computer Science is my secret lover. I initially studied that, but my university focused on producing java devs for industry and didn't do the "science" of computer science. I still work through books when I have time though. About half-way with TAOCP V1.

I think this really depends but yeah I generally agree with you - with respect to machine learning as an industry - but there is still fairly cutting edge ML being done at large companies which can use a lot of intense math - though as you mentioned that does generally require at least a graduate degree of some sort.

> climate science

Since you brought it up, I've had a hard time looking at positions for software engineers (not a single mention on this month's whos hiring for example). Any ideas?

Personally, I specialized in cryptography, did a few years of research in security, and now work in security. Note that I always liked programming (in fact I kind of chose math over eng pretty much randomly). To your point, I don't do a lot of math on a daily basis -- I see it more like painting/singing, you keep it as a passion, you don't do just that for living.

Domain experts that can code are among the highest paid people in several industries. So, if you don't know how to code, it would likely be a valuable skill to learn. Python is probably sufficient for many industries, though some of the performance-minded fields want C++ and even Fortran(!).

Lots of scientific computing requires a lot of math and is used in many industries (a company I did contract work for in the past had contracts doing fluid dynamics for P&G, geological analysis for ConocoPhillips, and something for NASA...all using the same set of technical tools, specifically Python, NumPy/SciPy, and some C++/Fortran backend stuff for performance; the company was founded by a math PhD and they employed multiple math PhDs and at least half of the people they employed had some sort of significant math background).

So, if I were in your shoes, I'd learn SciPy/NumPy, and start following the related communities for job postings. There's a lot of demand and a lot of good-paying positions surrounding those communities.

Data science works with math, too, but not often at the advanced levels it sounds like you might be looking for. AI/ML is also somewhat mathy, but not as much as one might think (though I get the feeling there's room for more mathy approaches to problems that are currently very brute-forcey, but my understanding of math and of AI/ML is low enough to where that gut feeling could very well be wrong).

Finally, digital signal processing is an area with significant reliance on math (though pretty specific sub-genres of math). I've recently started taking a couple of math MOOCs to refresh my memory of higher maths (and to learn it for the first time in some cases) so I can understand DSP algorithms for audio and music a little better. DSP has many applications in mobile devices, voice recognition, music and audio and video, etc.

If you're interested in biomedical sciences and curing diseases, then a career in bioinformatics/computational biology might be worth considering: http://www.bioinformaticscareerguide.com/p/career-guide.html

I used to manage a data science team for an ad-tech company. Applied modeling problems come up a lot in advertising - how to model an event as a probability distribution, how to use several of these probability distributions to solve optimization problems in order to get most benefit given a fixed budget, and so on. During my time managing that team, we had a few former PhD's (both graduated and all-but-dissertation).

Since you asked, I currently work at Amazon, and we're hiring PhD's as applied scientists: https://www.amazon.jobs/en/search?base_query=phd+mathematics

HPC companies hire many mathematicians. I work with several PhDs in applied mathematics.

Another area is finance, as you mentioned. I interviewed at several hedge funds and investment banks after my undergraduate degree (in physics and mathematics).

Of course, it's not hard to go into CS for a masters or PhD which opens up many other options as well. This was more what I did.

Almost all these positions require you to be capable of both whiteboard work, along with programming. So be sure you can develop software.

When statisticians rebranded themselves as data scientists a few years ago they got themselves a pretty big average salary bump for their trouble.

I worked at a company that made CAD/CAM software and had to make extensive use of calculus / linear algebra / computational geometry. Depending on what kind of math you like doing / using, that might be an industry worth looking into. Though it was a software development job, so I'm not sure if that aligns with your interests either.

My colleague has a 1st class honours degree in Maths from Cambridge university, he is both a mathematician and a software engineer. We work on radar tracking algorithms, for civil and military aircraft in the UK. Developing and/or understanding such algorithms needs highly skilled people who understand Maths; I consider them to be Mathematicians.

Professional societies like SIAM have listings, which you can view to get a rough idea: http://www.siam.org/careers/ . Most of these are academic, but not all. (I think, or at least hope, there's a way to filter/sort.) Most of them may be targeted at PhDs, but again not all.

I'm an academic myself, and a significant fraction of our recent PhD grads have gotten positions in data science, and at least one is now at a pharmaceutical company working on mathematical modeling. The jobs are out there. The tricky thing is that people don't generally advertise for mathematicians, even though a good mathematician may fit the job well.

Insurance companies. Look into the actuarial profession.

I came here to say this. From what I gather, actuaries make big money and have significant power at insurance companies. (I was merely a gumby processing claims and needed special training to do my entry level job.)


I've co-founded three companies (www.futurescaper.com, www.podaris.com, and www.imatest.com, to a much lesser extent), all of which have employed mathematicians on at least a part-time basis. If you're doing fancy parametric constraints-based modelling systems, analysis of n-dimensional causal graphs, or analysis of digital imaging systems, then a robust mathematics background can really help. Non of our hires have been been pure mathematicians, inasmuch as we've needed them to also be able to code. But I can easily imagine that larger companies pursuing similar domains ought to see the value in having purer mathematicians on their payroll.

One piece of advice I would give you is to think about where your particular area of focus in math overlaps with CS, how you could apply that towards a real world problem / situation and how you could tailor your education towards that goal. People in Math / ML / Stats with the chops to translate their research into production quality code and systems are in huge demand. PS, we are hiring data scientists at Wheelhouse! https://boards.greenhouse.io/wheelhouse/jobs/811845

You're doing it the wrong way round. Who needs mathematicians? What can it be used for? Find the appropriate people on forums, github or whatever and talk to them about it. Not HR, people who are involved in whatever you want to be involved in.

Maybe get to know this too: https://www.jetbrains.com/pycharm/features/scientific_tools....

The NSA is allegedly the biggest employer of mathematicians in the US, and possibly the world -- which is troubling, considering how ethically questionable their work is.

As opposed to say working to develop the software for those poker machines aka the crack cocaine of gambling. Even my (Scary)Great Great Uncle who was an off course book maker pre WW2 in Birmingham might blanch at that.

And yes it does make watching peaky blinders interesting :-)

I know several top-level mathematicians (International Math Olympiad medalists) and not all of them are working as Quants. Some of them compose music, some of them are working as machine learning experts, VC funds, etc. They may be slight outliers but it's possible to do whatever you want as long as you posses the intellect and drive. I think Mathematics definitely help with developing your logic skills that are applicable to anything.

My friend who did a math BS + partial masters from a large state agricultural university got a job at a company similar to this http://www.tricore.org/. understanding statistics and working with VBA in excel were important, moreso than programming which looked like the easy part. There were a couple other people with PHD in math who also work there.

Data engineering and data science teams at Avant are biased towards hiring pure math graduates. I think it's because training in mathematics necessarily teaches you to jump up and down between different layers of abstraction which makes motivated mathematicians very quick and effective learners.

If you don't mind relocating to Chicago drop me a line at kirill.sevastyanenko (at) avant

I did a BSc in math/econ, then MSc in math. Went into consulting work, mostly in security. Now work in crypto for a tech company.

We just made a maths-heavy hire where a majority of work is expected to be machine learning, linear algebra, forecasting, data analysis, etc. We design food preparation and retail robots and supporting logistics and automation systems.

Previously I tried to hire a maths PhD out of academia but the guy was all talk and no action. He's now working for a hedge fund.

I recently got a PhD in pure mathematics, and also recently started my job as a junior developer in a large consultant firm in Norway. I only knew basic Java and Python, but coming from mathematics, I find learning formal stuff quite easy.

I've been working now for 4 months, and I really enjoy it. I'm still learning something every day.

I'm at NIPS (#NIPS2017) right now and both computer science and industry need more people with strong math backgrounds. Advice: learn how to program and you'll never be in need of interesting-enough mathematics


Learn functional programming. It feels like math to a math person. Imperative programming feels very foreign (e.g., standard python).

Formal specifications and verification, proof assistants; program analysis and synthesis; cryptography from protocols to at-rest; compilers and performance analysis, etc.

I view maths as a power-augmenting skill set, like programming (which is a kind of applied maths). It lets you think big thoughts.

It also lets you break down the field into discrete chunks. A proof is not a field of study, but a field of study (in Mathematics) cannot exist without proofs. The way to start a field of study is to write a discrete conjecture which could be developed into a proof.

Software is the same way. All big programs had to start somewhere.

Look up "data science" positions wide range of applications and needs depending on the company

What are your interests? What kind of work/industry are you looking for?

I graduated MMath in 2012 and have just completed my 5th year as a Software Engineer at Bloomberg. I had a lot of programming experience, which helped. We have a specialist graduate program for non-CS grads.

Quant desks in hedge funds and prop trading companies - like Jane Street, SIG and the like.

In most of those places, 'maths' will involve a lot of data science and machine learning, plus a fair bit of programming, however PhD degree is quite often a hard requirement.

(Source: am a quant)

Please also look at genomics/bioinformatics. Lot's of mathematics + software involved.

US federal government is looking for a few: https://www.usajobs.gov/GetJob/ViewDetails/478727900

GCHQ / CESG / whatever the equivalent is in your country.

See also arms length organisations, such as Heilbronn Institute. https://heilbronn.ac.uk/

We have a large number of MS/PhD level Math folks in my Data Science group.

Actuarial work is quite maths-y and grad schemes typically want maths grads as they'll have the requisite grounding to do the actual work. It is not fully finance but it is obviously money/risk related.

Totally anecdotally, I've had friends with undergraduate degrees in Math go straight into consulting and software engineering out of college with limited work experience.

Neptune technology group. Water meter company since 1892. They started making ultrasonic meters a few years back and are looking for an applied mathematician.

Mathematicians can be excellent programmers, especially in machine learning and statistical fields. I am a programmer and would love to be a mathematician.

I graduated with an MS in Applied Mathematics and ended up working as a data scientist for 4 years until moving on to more interesting stuff.

What interesting stuff did you move onto?

web consulting, i miss the math though, dont use it too much anymore.

The NSA and specialized consulting firms in the Washington, DC area ( Reston )

Aranz Geo/ Seequent does providing you can code to

Who hires physicists? Asking for a friend.

IIRC Wall Street is the biggest employer of physicists.

They get all the really nutty math and can apply it to models, etc.

That might just be Ph. D's though.

Is that still true? Now that exotic securities are looked down on, and the safe ones are established don't require Ph. D level research anymore?

Follow up: You are right of course about Wall Street. NY, there's no better place for a physicist that is tired of being broke. But, is the same true about other big cities? Chicago, LA, Houston, SF? Is there any chance? Or is NY Google, and Chicago is Bing at best?

Have you considered the NSA or CCR?

Pretty much anything. Any consultancy would probably hire you as an analyst for example if you social skills are ok.

Casino game publishers in Vegas

banks, small armies of them...

NSA is great for mathematicians, but competition is tough. The best way to get in as a co-op early in your college years, since it takes a year to get a clearance.

really that long? I though the NSA / CIA would expedite clearance.

So if you want to intern at the NSA do you apply a year in advance? maybe Matt Cutts knows

Yes. It really takes around 9 months. You'd have to apply beginning of your sophomore year.

Spend a couple of months studying deep learning and python and you will get a very well paid job.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact