I will describe the goals and strategy of the SWARMS survey, a data mining project designed to find the double degenerate progenitors of Type Ia Supernovae. SWARMS takes advantage of the multiple spectra of thousands of white dwarfs collected by the Sloan Digital Sky Survey. The first results include the two detached white dwarf binaries with the shortest known periods and a mysterious binary with a combined mass that could be very close to the Chandrasekhar limit. I will examine the implications of these discoveries, discuss the potential for future developments, and examine the importance of data mining techniques for forthcoming projects like LSST and Gaia.