We introduce BayeSQP, a novel black-box optimization algorithm that combines sequential quadratic programming with Bayesian optimization for high-dimensional constrained problems..