EXTENDED ABSTRACT: I will show how a relatively simple phenomenological model encoding the physics of phase transitions can make a prediction that guides data science towards ffnding a new alloy. The problem I will address is the design of superelasticity in TiNi based alloys over a wide temperature range. Many applications require materials capable of operating under extreme conditions. For example, the Mars probe must withstand the rough Martian terrain within a broad temperature range from -140 C to 150 C. I will show how a simpliffed Landau theory can provide a prescription for the nature of the phase boundary that will lead to a wide temperature window. Data Science then allows us to rapidly search for the alloy with the given phase boundary.