It's not late to switch at 36. My suggestions would be * do not throw your past experience and start from zero, cash on your experience: learn some Data Scienc

It's not late to switch at 36. My suggestions would be * do not throw your past experience and start from zero, cash on your experience: learn some Data Science / ML / Data Engineering and transition to the role where you can work closer with them: building data services, productionising ML, etc. * try before committing. I personally thought that due to my experience in software and passion for mathematics ML would be a natural chemistry for me, but I found it quite boring (comparing to both doing pure mathematics and building software). * the area is still hyped and there are lots of people who wants to get there - it raises the hiring bar.