The structure and rheology of randomly branched polyamide melts, in particular that of branched polyamide 6, are predicted on the basis of their initial reaction recipe. To this end, a Monte Carlo approach has been developed in order to build different molecular architectures from the initial reactant monomers at the appropriate conversion level. This approach allows us to analyze the composition of these melts in terms of topological architecture and molecular weight of the various polymer species present. Subsequently, the linear rheology of each sample is predicted within the tube model framework [van Ruymbeke et al., Macromol. 2006], based on the position/seniority of the different branches in these polymer species, by averaging over a limited number of representative segments. This approach allows us to discuss the role the different polymer architectures play in the overall viscoelastic response, the importance of the different initial monomers to adequately tune the composition (and thus the rheology) of these branched systems and the necessity of reaching a high conversion level in order to obtain a large zero-shear viscosity. We also extend this approach by applying a bimodal distribution to describe the solid state polycondensation of these products. The predictions are in good agreement with experiments. This method may be applied to any branched polymer product that is synthesized via a melt condensation type reaction and can be used as a tool to test and screen in detail the flow properties of materials without prior synthesis.