It's never too late. Maybe apart from some age bias in the field (but that most occurs closer to 40). I would suggest starting with some good foundation in stat

It's never too late. Maybe apart from some age bias in the field (but that most occurs closer to 40). I would suggest starting with some good foundation in stats / modeling, this is an amazing book on that: An Introduction to Statistical Learning by Robert Tibshirani and Trevor Hastie.