Phage-therapy is a promising alternative against pathogenic, multiple drug resistant bacteria. In this work we propose an algorithm to determine the optimal bacteriophage dose able to minimize a population of Pseudomonas aeruginosa. Reverse engineering was used to determine the kinetic parameters; subsequently, a bi-level optimization platform was implemented for a model based on evolutionary programming. Our prediction of optimal dose was tested in vitro with planktonic cultures of P. aeruginosa . From the data obtained, we conclude that reverse engineering and stochastic simulations are a useful approach to find optimal phage doses against pathogenic bacteria, an important step for the implementation of phage-therapy.