Machine translation systems are a necessity in today’s globalized world. Statistical and neural approaches to machine translation allow the construction of these systems automatically from bilingual and monolingual datasets. The Machine Translation group has the capacity to design the processes necessary for the acquisition and preparation of the data sets necessary for the process of learning the translation models, as well as the experience in the search for the best configuration of these models. This group also has experience in the use and fine-tuning of pre-trained multilingual language models for the construction of machine translators with a reduced production cost