Andrea Santos Revilla
PhD Candidate in Neuroscience & AI
I am a computational linguist with an interdisciplinary background encompassing informatics, language, and neuroscience. My PhD project focuses on speech production decoding in the child brain, where I apply machine learning models to iEEG neural data collected from patients undergoing treatment for drug-resistant epilepsy. The goal of this project is to better understand the neural mechanisms of language production across the developmental age. Before joining the lab, I completed an MSc in Computational Linguistics and Neuroinformatics at Universität Zürich and ETH Zürich, following a BA in Linguistics from University College London.
My technical expertise sits at the intersection of speech processing and machine learning, with research experience in automatic speech recognition under atypical situations, such as non-native children's speech or pathological speech. I have also worked on evaluating the language learning trajectory of language models, with a specific focus on prosody, and on diffusion models for EEG data.
Over the next five years, my scientific interest is focused on modeling typical language development in children to establish a stronger baseline for researching atypical development and understanding the specific impact of drug-resistant epilepsy on a child's linguistic growth.
