First-principles calculations (DFT, DMFT) are very successful in describing materials at a unit cell scale. Simplified models have often inspired these approaches and have a great pedagogical value, as well as enable modelling at larger scales, where the performance of functional materials is often defined. For instance, switching of ferroic materials is determined by the dynamics of their complex domain arrangements, and novel electronic and memory devices are being developed using domain walls and skyrmions in non-collinear magnets and multiferroics. Recent developments of Wannier function-based methods, including computation of topological invariants and electoron phonon coupling, as well as density functional perturbation theory approaches for treating gradients of strains and other order parameters, enable the computation of coefficients of complex models. Methods for automatic determination of model parameters from DFT are being developed. Notably Scale-Up allows to automatically generate and simulate an atomistic model representing structural distortions, strains and selected electronic degrees of freedom. Multibinit project aims to bridge Abinit DFT code and atomistic models. L-Invariant uses symmetry to derive a Landau-type model and performs automatic generation of DFT data set, subsequent fitting of the model parameters and Monte-Carlo and phase-field simulations of the models. The proposed school aims to provide the students with the review of important phenomena in solid state physics, modeling approaches, ab-initio methodology and practical sessions that combine first-principles and model calculations on realistic materials. Topic will include elastic and ferroelectric properties, magnetic exchanges, particularly in materials with strong spin-orbit coupling, interactions between (uniform and nonuniform) strains, structural and magnetic order parameters, finite-temperature properties, basics of strongly correlated electrons and electron-phonon interactions.