Natural Language Processing (NLP) has revolutionized the automatic reading and processing of texts, culminating in the emergence of Large Language Models that have made it possible to understand and extract information at unprecedented levels. This study explored the use of NLP and Machine Learning techniques to segment UCITS (Undertakings for Collective Investment in Transferable Securities) investment fund prospectuses in order to optimize the collection of unstructured data contained therein. Using n-gram frequency and clustering algorithms, this thesis seeks to expand the application horizons of Artificial Intelligence in the scope of financial markets.