There are 3 steps in a Bayesian workflow. The first where avaialble knowledge is captured about a given parameter in a given statistical model through prior distribution. A likelihood function which is determined using the information about parameters found in the observable data and posterior distribution.