All depends on the type of math you specialized in and enjoy.
There are great opportunities in the Big Data side - not as a data scientist (although there is opportunity there) but on the algorithm development side. Basically there are a plethora of algorithms developed in academia which have not yet (and maybe never will be) imported into the machine learning/Big Data space. The maths aspect is both to convert these algorithms into cloud capable scalability - which is a computer science type endeavor - as well as the automation of the parameterization - which (in my limited understanding view) seems like a much more maths oriented need. I'd look at companies like Mahout, Spark, H20, etc to see what opportunities might be there such that you could learn about that space before heading out to seek your own strike in algo mining.