NLPstudio
Scalable, reproducible NLP workflows in R — corpus preparation, dictionaries, readability, similarity, and a consistent topic-model API across backends. Built on…
I am an Assistant Professor of Accounting Analytics and Data Science at the Department of Accounting at Bocconi University, a Fellow at the Bocconi Institute for Data Science and Analytics, and a member of the Bocconi Digital Transformation and AI Committee.
I study core questions in accounting and finance — corporate disclosure and its consequences for firm tax outcomes, investor reactions to multi-dimensional financial narratives, and the dynamics through which corporate information reaches the market — using Natural Language Processing, machine learning, and computational statistics on large-scale textual and multimodal financial data. Methodologically, I draw on statistical learning and stochastic processes for sequential data, which let me model how corporate information evolves over time (disclosure cascades, topic model identification, and more).
A second strand of my work is deliberately cross-disciplinary, cutting across accounting, marketing, statistics, and computer science. There I treat artificial intelligence and generative AI as objects of study rather than instruments: how large language models behave when asked to measure, judge, or stand in for people, and what researchers can legitimately expect of them — whether general-purpose models rival domain-calibrated ones, how much of a measure survives the numerical details of the computation behind it, and whether reproducing canonical behavioral effects is enough to call a model human-like. The same methodological impulse runs through my work on diagnostics for topic models, which brings regression-style effect sizes and residual checks to unsupervised text models.
My work has appeared in the Review of Accounting Studies, the Journal of Machine Learning Research, the Journal of Accounting and Public Policy, and other outlets.