In my experience there is a lot of demand for data scientists (e.g., if you look at freelance small job sites like Upwork), but there are also lots of data scientists, who actually did their degree in data science. It is usually difficult to compete with them, if your background is in another field.
On the other hand, I found that there is demand for good math skills with moderate requirements for programming when it comes to developing algorithms for analyzing domain-specific data: modeling for biological systems, counting molecules/proteins, climate science, modeling of various kinds of sales, breaks, flows, etc.
I worked on a similar kind of project where I had to implement an algorithm from a scientific paper in python or R, and then pass it to professional developers for writing an industrial scale code. Another type of project involves evaluating feasibility of certain type of algorithm.
However, besides math and moderate coding skills, this requires some domain knowledge and reasonable background in statistics.