In this paper, we present a method to solve the quantum marginal problem for symmetric (d)-level systems. The method is built upon an efficient semi-definite program that determines the compatibility conditions of an (m)-body reduced density with a global (n)-body density matrix supported on the symmetric space. We illustrate the applicability of the method in central quantum information problems with several exemplary case studies. Namely, (i) a fast variational ansatz to optimize local Hamiltonians over symmetric states, (ii) a method to optimize symmetric, few-body Bell operators over symmetric states and (iii) a set of sufficient conditions to determine which symmetric states cannot be self-tested from few-body observables. As a by-product of our findings, we also provide a generic, analytical correspondence between arbitrary superpositions of (n)-qubit Dicke states and translationally-invariant diagonal matrix product states of bond dimension (n).