Geoffroy Dubourg Felonneau
About Geoffroy Dubourg Felonneau
Geoffroy Dubourg Felonneau serves as the Director of Machine Learning at Shiru, where he leads the development of machine learning models for protein discovery. He has a background in machine learning and engineering, with previous roles at Cambridge Cancer Genomics and Thales.
Work at Shiru
Geoffroy Dubourg Felonneau serves as the Director of Machine Learning at Shiru, a position he has held since 2024. In this role, he guides the development and training of innovative machine learning models focused on protein discovery. He designs development processes aimed at uncovering sustainable protein-based solutions for various industries. Additionally, he architects scalable infrastructures for machine learning applications, enhancing the overall efficiency and effectiveness of the team's projects.
Previous Experience in Machine Learning
Prior to his current role, Geoffroy worked at Shiru as the Machine Learning Lead from 2020 to 2024. His experience also includes serving as the Machine Learning Team Lead at Cambridge Cancer Genomics from 2018 to 2020. He began his tenure at Cambridge Cancer Genomics as a Machine Learning Engineer from 2017 to 2018. His diverse roles have contributed to his expertise in machine learning applications within the life sciences sector.
Education and Expertise
Geoffroy Dubourg Felonneau earned a Master of Science (M.Sc.) in Computer Science from CY Tech, where he studied from 2010 to 2015. His educational background provides a strong foundation in computational methods and algorithms, which he applies in his current work in machine learning and protein discovery. His academic training supports his innovative approaches in developing AI methods for predicting protein functions.
Research Contributions
Geoffroy has made significant contributions to the field of machine learning and protein research. He is the lead author of a paper presented at NeurIPS, a prominent AI research conference. His work includes creating a sequence-to-function map of the protein universe and developing a new AI method for predicting protein subcellular localization. He has combined Language Models and Graph Neural Networks to enhance protein function prediction, contributing to advancements that outperform existing state-of-the-art methods in protein localization prediction.
Early Career Development
Geoffroy's early career included an internship at Thales in 2014, where he gained practical experience in engineering for five months. This role provided him with foundational skills that he later applied in his machine learning career. His progression from an intern to leadership roles in machine learning demonstrates his commitment to professional growth and expertise in the field.