Lorena Calvo-Bartolomé is a Postdoctoral Research Fellow at Bocconi University. Her research broadly focuses on making complex models interpretable and useful, that is, understanding what they learn, and turning those insights into tools that real people can act on. During her PhD, she focused on the industrialization of topic models, designing systems that let domain experts guide, evaluate, and deploy models at scale for decision-making. She now extends this interpretability lens to large language models, exploring how their internal representations can be understood and leveraged, much like reading the topics of a topic model, but at a deeper level of abstraction. She is also keen on designing task-oriented evaluation protocols and interactive systems that put domain experts at the center of the modeling process. Lately, she has been exploring voice-based data collection as a richer alternative to written annotation in NLP, combining her signal processing roots with her current work.