Computational design of magnetic motors, generators, and sensors is common practice at the device level: Finite element software is used to optimise the geometry of magnetic motors. Sensor models are used in the software tools for electronic circuit layout. The functional properties at the device level depend on the magnetic properties of the materials that in turn arise from the interplay between the quantum physics electronic properties at the atomistic length scale and the material’s nanoscale and microscale structure. In order to address these vastly different length scales independently, there exist software tools that are mainly developed in academia. These disconnected tools are based on (i) density functional theory simulations to compute intrinsic magnetic properties, (ii) Heisenberg (also known as spin dynamics) models to study collective and stochastic behaviour of thousands to millions of atoms, and (iii) micromagnetic simulations using finite difference or finite element methods to compute the impact of shape and internal structure on the hysteresis properties of magnetic materials. MaMMoS will be built on this existing know-how and prepare a software suite for computational materials design for widespread use in European industry and research. MaMMoS will build data-driven models for magnetic materials at lengths ranging from the atomistic scale to device level through the fusion of characterisation and modelling data. MaMMoS aims to link the existing software tools across the different length scales, to establish standards for data exchange in applied magnetism, and tailor the software suite according to European industry needs.