At the heart of numerous applications, especially those related to energy generation, storage, and its efficient use, lie semiconductors, with their atomic-scale electron behavior being a common crucial factor. The primary aim of this project is to deeply understand the complex physical phenomena in semiconductors that impact device efficiency and to control and optimize them using electronic structure and atomistic modeling. This involves tackling issues such as charge trapping, defect formation, and the influence of atomic dynamics on electronic structures, as well as the intricate properties of mixing, surfaces, and interfaces alongside band alignments. Our holistic approach aims to improve semiconductor devices such as solar cells based on halide perovskites, photoelectrochemical cells using oxides and oxyhalides, and tunnel field effect transistors involving heterostructures of transition metal dichalcogenides (TMDs). We will perform simulations of electronic structure and absorption properties using codes like VASP, CP2K, or ABINIT, as well as molecular dynamic simulations employing neuroevolution potentials implemented within the GPUMD code. This will provide us with a comprehensive toolkit for the optimization of advanced semiconductor devices.