Andrei Stefan Bejgu

AI Applied Scientist @ SylloTips | Ph.D. in Artificial Intelligence @ Sapienza University of Rome

Andrei Stefan Bejgu - AI Applied Scientist specializing in NLP, LLMs, and multilingual AI

👋 Hi! I’m an AI Applied Scientist at SylloTips in Rome, where I build production systems that bridge academic research and real-world applications. My work focuses on developing human-in-the-loop AI agents, retrieval-augmented generation systems, and fine-tuning multilingual LLMs at scale.

🎓 Research & Engineering

I earned my Ph.D. in Artificial Intelligence at Sapienza University of Rome, specializing in multilingual NLP and advanced machine learning. My industrial Ph.D. was conducted in collaboration with Babelscape, where I gained hands-on experience turning research into production systems—prioritizing both theoretical rigor and practical deployment.

Current Areas of Focus:

Human-in-the-loop AI Agents RAG Architectures Multilingual LLMs Word Sense Disambiguation Semantic Understanding Distributed Training

Beanis — Redis ODM for Python

A high-performance Redis Object-Document Mapper that reduces boilerplate by 70% while maintaining performance within 8% of raw Redis. Built for developers who need type safety without sacrificing speed.

Word Sense Linking — ACL 2024

Research on automatic disambiguation that works on real text, not just controlled examples. Tackles the challenge of understanding context-dependent word meanings at scale.

Production RAG Systems

Building retrieval-augmented generation pipelines that developers can actually deploy and maintain in production environments with real-world constraints.

💡 I enjoy collaborating with cross-functional teams and mentoring engineers at all career stages. When I’m not writing code or papers, I’m usually exploring new ideas in AI or sharing what I’m learning through blog posts and open-source contributions.

latest posts

selected publications

  1. CL
    Truth or Mirage? Towards End-to-End Factuality Evaluation with LLM-OASIS
    Alessandro Scirè*, Andrei Stefan Bejgu*, Simone Tedeschi, and 3 more authors
    Computational Linguistics, 2025
  2. EMNLP
    Concept-pedia: A Wide-coverage Semantically-annotated Multimodal Dataset
    Karim Ghonim, Andrei Stefan Bejgu, Alberte Fernández-Castro, and 1 more author
    In EMNLP, 2025
  3. ACL
    Word Sense Linking: Disambiguating Outside the Sandbox
    Andrei Stefan Bejgu, Edoardo Barba, Luigi Procopio, and 2 more authors
    In ACL, 2024
  4. CroCoAlign: A Cross-Lingual, Context-Aware and Fully-Neural Sentence Alignment System for Long Texts
    Francesco Molfese, Andrei Stefan Bejgu, Simone Tedeschi, and 2 more authors
    In EACL, 2024