General information

Full Name Pietro Lesci
Languages Italian, English


  • 2021 - Present
    PhD in Computer Science
    University of Cambridge
  • 2016 - 2019
    MSc in Economic and Social Sciences
    Bocconi University
    • Dissertation: "Deep Learning: A Statistical Perspective"
  • 2013 - 2016
    BSc in Economics and Management
    Università del Sacro Cuore
    • Dissertation: "Market Sentiment and Monetary Policy"
  • 2008 - 2013
    Scientific High School Diploma
    Liceo Scientifico "Enrico Mattei" (Castrovillari)
    • Dissertation: "Market Sentiment and Monetary Policy"

Research experience

  • Present
    Applied Scientist intern
    Amazon AWS - AI Labs
  • 2022
    NLP Engineer
    Theia Insights
    • Development of NER annotation pipeline
    • Development and training of transformer-based models for NER
  • 2019 - 2020
    Research Assistant
    Bocconi Institute for Data Science and Analytics
    • Working with Prof Dirk Hovy, as part of the DMI research unit, on the identification of echo chambers and online abuse on Twitter
    • Development of machine learning web-apps: Wordify and MACE


  • 2020 - 2021
    Senior Associate, Data Science
    Bain & Company
    • Data-driven quality assurance (industry: mining)
    • Pricing optimization (industry: packaging machines)
    • Predictive maintenance in hydropower plants (industry: energy)
    • Client segmentation (industry: road tolls and telco)
    • Development of an internal Python library for NLP
    • Deployment of NLP end-to-end solutions for product categorization
  • 2021
    Software Engineer (part-time contractor)
    HDM Group
    • Development of a time-sheet management tool (industry: energy)
  • 2018
    Data Science Trainee
    European Central Bank
    • Data quality management of private and public sector granular master data for the Directorate of General Statistics
    • Database reconciliation: lead of the RIAD-GLEIF project
    • Selected member oftheRIAD–AnaCredittask force with experts from the National Central Banks and the financial sector
    • Speaker during regular plenary meetings with experts from the National Central Banks (reserved only to a restricted group of trainees)
    • Automation of a data ingestion and deduplication procedure (80% efficiency gain) used in the computation of the Euro Short-Term rate
    • Development of automated internal reports and dashboards
    • Production of statistical reports published on the ECB website

Open-source projects

  • 2022 - Present
    • An active learning library for PyTorch
  • 2019 - 2022
    • Wordify makes it easy to identify words that discriminate categories in textual data.
  • 2019 - 2022
    • MACE makes it easy to compute inter-annotators agreement statistics.

Reviewer experience

  • EMNLP 2022
  • ACL 2021

Other interests

  • Enrolled in the BSc in Jazz, GPA 29.7/30
  • Classical Piano classes (12 years)
  • Bass Guitar classes (10 years)
  • Sol-Fa degree and Piano and 5th year diploma
  • Volunteering at Associazione “Il Sorriso” ONLUS
  • Organist and Choir Director
  • Singer, bassist, and arranger in jazz ensembles and a rock trio