What jobs in Mathematics are always in demand, and are deeply Mathematically specialised or greatly general? I am wondering what jobs in the field of Mathematics are (seemingly) always in demand. I am also wondering what jobs there are that are (once again seemingly) greatly Mathematically demanding in regards to either deep specialisation or great generality  in the application fields.
For the second criteria, I mean that a great amount of learning must be done to understand the career. Whether that be learning a lot of content from multiple fields, or significant investment in a specialization.
More specifically, what fields meet both of these criteria?
Thank you for your time.
Note: If you wish, give it in regards to any location you wish, I understand this can be (and likely is) location reliant.
 A: With the arrive of big data, Statistics, as a branch of Mathematics, is now in high demand. And the challenges there are in many areas. Buzz words like Machine Learning, Analytics, Computer Vision and so on have statistics at their core.
Personally, being a software engineer with a maths background, I recently engaged in postgraduate studies in statistics, and I can say that I've been headhunted couple of times since then. Lucky, my current position involves both, software engineering and statistics; but the interest from companies (mainly the big ones) is there!
A: With regard to your first criterion, Actuaries are often in demand.  That's not to say that there aren't Actuaries who are out of work.  The "problem" with being an Actuary is that you've got to pass their series of 9 (or 10) exams, where not all of them are mathematically oriented.  You can be an Associate Actuary by passing 7 exams, but that wouldn't be as advantageous as passing all of them.
With regard to you second criterion, the work may or may not be mathematically challenging, depending upon what you wind up doing.  There are definitely mathematical challenges, but one could still be a highly paid Actuary and spend more of your time in the business end of things than the highly technical end of things.
