To date, the vast majority of work on metal AM has been framed in terms of the need to tune processing conditions for the AM of X alloys, where X stands for conventional alloy classes (e.g. SS316L, IN718, Ti64, etc). This approach often overlooks the fact that historically, every engineering alloy has been designed with a processing route in mind. The question then is how best to approach the alloy design problem from a perspective that incorporates notions of printability or processability. In this talk, I will introduce our efforts to tackle the alloy design problem while explicitly considering alloy features associated with the suitability of a given alloy to be printed using laser powder bed fusion (LPBF). Our framework combines physics-based models, CALPHAD phase stability predictions, machine learning, Bayesian optimal experimental design. We apply our framework on the discovery of potentially printable refractory high entropy alloys.