About me

I am a final-year Ph.D. student in Computer Science at Politecnico di Milano, where I work under the supervision of Prof. Paolo Cremonesi. My research focuses on developing and evaluating AI agents that can safely and effectively interact with the real world.

Research Interests

My research lies at the intersection of AI Agents, Multi-Armed Bandits, and Large Language Models. I am particularly interested in:

• Decision-making with Large Language Models
• Uncertainty estimation in Large Language Models
• Off-Policy Evaluation and Learning for AI agents
• Applications to recommendation systems and personalization

Selected Publications

• On the Importance of Uncertainty in Decision-Making with Large Language Models
  TMLR 2024
  with Lucas Maystre, Sina Ghiassian, Kamil Ciosek

• Off-Policy Evaluation with Deficient Support Using Side Information
  NeurIPS 2022
  with Maurizio Ferrari Dacrema, Marcello Restelli, Paolo Cremonesi

• Automated Off-Policy Estimator Selection via Supervised Learning
  CONSEQUENCES Workshop @ RecSys 2024
  with Michael Benigni, Maurizio Ferrari Dacrema

Experience

• Applied Scientist Intern @ Amazon (Summer 2024)
  Greater Los Angeles Area, CA, US
  Worked on large-scale machine learning applications for personalization

• Research Scientist Intern @ Spotify Research (Summer 2023)
  London, UK (Remote from Italy)
  Researched uncertainty estimation in LLM-based decision-making

Education

• Ph.D. in Computer Science, Politecnico di Milano, 2021-2025 (expected)
• M.Sc. in Computer Science and Engineering, Politecnico di Milano, 2018-2020
• B.Sc. in Computer Science and Engineering, Politecnico di Milano, 2015-2018

Contact

Feel free to reach out for research collaborations or discussions:
nicolo DOT felicioni AT polimi DOT it