This talk introduces a novel control-theoretic framework for global non-convex optimization. We begin by reformulating the minimization problem as a discounted infinite horizon optimal control problem, whose associated Hamilton–Jacobi–Bellman (HJB) equation yields a value function that captures the full landscape of the objective. By extracting gradient information from this value function, we obtain a feedback control that drives trajectories toward the global minimizers of general non-convex functions. Theoretical results establish explicit convergence rates for the variational behavior of the value function, the objective function value, and the pathwise convergence of the resulting trajectories to the minimizer set under minimal structural assumptions that require neither convexity, differentiability, nor Lojasiewicz-type conditions on the objective. Furthermore, the HJB-derived control can be integrated into the Consensus Based Optimization (CBO) framework for improved scalability. Numerical experiments demonstrate the effectiveness of the approach on challenging non-convex benchmark problems.