The standard approach for treating the electron-phonon coupling from first principles is density functional theory within so-called semi-local approximations (LDA and GGAs). These methods fail, however, dramatically in systems with localised d or f electrons, failing to describe their insulating phases as well as their superconducting properties. Recent works have shown that improving the description of electronic correlation beyond semi-local approximations can strongly enhance the electron-phonon coupling, substantially increasing the critical temperature of certain materials exhibiting unconventional superconductivity. In this project we will investigate why semi-local approximations fail and how to overcome their deficiencies with the design of fully nonlocal approximations able to simultaneously capture short and long range effects of the Coulomb interaction. As an application, we will study TiSe2, an intriguing d electron system where superconductivity competes with a charge density wave instability.