There is an urgent need in the field of soft tissue engineering (STE) to develop biomaterials exhibiting a high degree of biological and mechanical functionality as well as modularity so that they can be tailored according to patient-specific requirements. Recently, biomimetic soft network composites (SNC) consisting of a water-swollen hydrogel matrix and a reinforcing fibrous network fabricated by melt electrospinning writing technology have demonstrated exceptional mechanical and biological properties, thus becoming strong candidates for STE applications. However, there is a lack of design approaches to tailor and optimize their properties in a non-empirical way. To address this challenge, we propose a numerical model-based approach for the rational design of patient-specific SNC for tissue engineering applications. The approach is rooted in an in silico design library that allows for the selection of biomaterial and architecture combinations for the target application, resulting in reduced time, manpower and costs. To demonstrate the validity of the design strategy, a multiphasic SNC with predefined zone-specific properties that captured the complex zonal mechanical and compositional features of articular cartilage was developed based on the natural design of the native tissue.