I program algorithms in Mathematica, to develop models while working very closely with the data.  Updated models are then programmed into Excel for day-to-day implementation.  If I were younger and trained specifically in stats, I’d probably use more R or Python.  I started studying AI in its early days, and for people interested in learning the skills to do all this, I suggest machine learning.

My “what going into the model” article gives a lot more detail about the parameters incorporated.