Machine Learning and Artificial Intelligence techniques are disrupting several fields of engineering, including telecommunications. In this domain, network operators are evaluating the use of data-oriented techniques to improve the management of virtualized networking infrastructures in several aspects, including monitoring, optimization, traffic forecasting, capacity planning, anomaly detection, and others. This paper summarizes our ongoing experience in the area of adopting human-machine interfaces based on Natural Language Processing for easing and speeding-up the interactions between these systems and network operators in their everyday work.