HEPfit is a framework for analyses of direct and indirect constrains on models in Particle Physics. Its statistical core relies on a Markov Chain Monte Carlo based Bayesian algorithm implemented in BAT. An important component of HEPfit is computing flavour observables to compare them against current experimental constraints and make future projections. I will highlight some past work done in flavour physics with HEPfit. I will also show how HEPfit can be used as a standalone library to implement a customized model do any kind of physics analysis with any model of your choice. The strength of HEPfit lies not only in what has already been encoded in the program but also in its flexibility to be used with new models and observables. This stands in contrast with what many other similar codes available today can do.