brms is a great package. It allows you to put predictors on a lot of things. Its power is however not absolute — one thing it doesn’t let you directly do is use data to predict variances of random/varying effects.