The Zorro project is located inside the Cerro Dragon field which is subdivided in nine blocks due to main faults delimitating the structure and the reservoir rock distribution. The majority part of the oil production of the project is associated to secondary recovery. This work presents the results of the application of a waterflood optimization workflow powered by data physics modeling using data from the Golfo San Jorge basin. The proposed workflow combines traditional reservoir engineering equations with machine learning, data assimilation techniques and advanced optimization algorithms. The models incorporate real data, which can be updated continuously, allowing the engineers to evaluate different scenarios and identify the optimum injection plans for the asset. As the models honors physical principles, this methodology provides long-term predictions and ensure physical feasible results. The result is a sample of the application of a dynamic workflow and how the digital transformation process is providing value to the oil and gas industry. The obtained production increase shows how multidisciplinary teams are enhanced by the inclusion of technology, managing to optimize processes and production.