Creator of the Relation Extraction model REBEL (400+ GitHub stars, 2M+ Hugging Face downloads)
Contributor to Minerva, the first family of Italian LLMs trained from scratch (1B, 3B, 7B)
Developer of the closed Information Extraction model ReLiK (250+ GitHub stars)
Teaching Assistant for M.Sc. Natural Language Processing course
Published papers at top-tier conferences including ACL, EMNLP, NAACL, and EACL
Research Experience
Postdoctoral Researcher at FAIR, Meta, Feb 2025 – Present; working on embeddings and embedding reasoning models under Holger Schwenk
Marie Skłodowska-Curie Ph.D. Fellow, Sapienza University and Babelscape, Nov 2020 – Nov 2024
Instep Intern at Infosys Bangalore, Jun 2019 – Aug 2020; developed a Transformer-based QA system for insurance documents; ranked #1 Best Internship 2020 by Vault
NLP Developer at Amsterdam Data Collective, Dec 2018 – Jun 2020; built scientific article recommendation systems and implemented Transformer models for client projects
Software Developer at BaseTIS, Feb 2017 – Aug 2018; web development with Angular; machine learning programming in Python and R