There are several expositions of D-modules, ranging from the extremely terse (Bernstein's lecture notes) to the very pedantic and thorough. But unlike all of these, this exposition uses almost no machinery (it is aimed at undergrads or first year grad students). On the one hand, that means it's not really an exposition of the theory: you can't define the direct image functor without derived categories. On the other, it's a helpful balance, and the concrete examples (such as adjoining the inverse of a polynomial to the polynomial ring to produce a holonomic D-module) are insightful and flow well.