Abstract: In this talk I will review some results on the mathematical modeling and control of social dynamics, like herding phenomena, crowd behaviors, traffic and opin- ion formation. In particular for an evacuation problem, I will consider the case where few controlled agents, possibly not recognizable by the multitude, want to enforce the emergence of a desired trajectory. In order to reduce the high complexity of these type problems, I will propose a gen- eral framework based on the approximation of the microscopic dynamic through a Boltzmann-like equation, showing that under a proper scaling this approach is equivalent to the mean-field description of the original model. Thereafter, I will present a set of fast algorithms, based on the simulation of the binary interactions of the Boltzmann equation, which allow to solve efficiently the control problem for the mean-field dynamic. Several numerical simulations will show the efficacy of the proposed strategies.