We introduce a novel harmonic analysis for functions defined on the vertices of a strongly connected directed graph (digraph) of which the random walk operator is the cornerstone. As a first step, we consider the set of eigenvectors of the random walk operator as a non-orthogonal Fourier-type basis for functions over digraphs. We find a frequency interpretation by linking the variation of the eigenvectors of the random walk operator obtained from their Dirichlet energy to the real part of their associated eigenvalues. From this Fourier basis, we can proceed further and build multi-scale analyses on digraphs. We propose both a redundant wavelet transform and a decimated wavelet transform as an extension of spectral graph wavelets and diffusion wavelets framework respectively for digraphs. The development of our harmonic analysis on digraphs thus leads us to consider both semi-supervised learning problems and signal modeling problems on graphs applied to digraphs highlighting the efficiency of our framework.