The sign problem in particle physics appears in QCD as soon as a non-zero chemical potential is studied. This prevents direct Lattice simulations to determine the phase structure of the strong force. Complex Langevin methods have been successfully used for various models or approximations of QCD, however, in some scenarios it converges to incorrect results. We will present a new method that helps to improve the convergence by staying close to the SU(3) manifold. Here we will focus on pure Yang-Mills simulations and QCD in the limit of heavy quarks, which provide a testing scenario for these simulations.