Future solutions in ultrahigh density magnetic recording, magnetic sensor technology and spintronics based logic devices require a deep understanding of the underlying physical mechanisms. Magnetic phenomena in solids are widely studied by Monte-Carlo or spin-dynamics simulations using spin models. Deriving the interaction parameters in these models from quantum mechanical (first principles) calculations increase the credence to these investigations. While changing the chemical composition in magnetic alloys is a sensitive tool to manipulate the interactions between the magnetic atoms, taking into account the local atomic configurations pose a particular challenge for the calculations. The main objective of the proposed PhD research is to simulate the short-range order (SRO) of the atomic constituents in the methods developed earlier in our group to calculate the magnetic interactions. To this end we will apply the Embedded Cluster Green's Function method within the Korringa-Kohn-Rostoker formalism. We will investigate the formation of local magnetic moments and the interactions between them in small ordered clusters embedded in chemically disordered host systems described in terms of the Coherent Potential Approximation, also available in our program package. We intend to study bulk magnetic alloys, like FeCo or FexMn3-xMn, the latter one showing intriguing magnetic phase transitions as a function of the Fe concentration. Beyond the bulk state, we will extend our studies to surfaces, interfaces and to multilayer systems that might display an enhancement of SRO-induced correlations due to their reduced dimensionality.