Senior Associate, Data Science

Bain & Company

Mar 2020 – Present Milano - Italy
  • Contributed to the development of a text analytics python library for topic modeling and sentiment analysis
  • Compiled a technical report on predictive maintenance in hydropower plants with machine learning methods
  • Customer segmentation (Microsoft SQL Server for ETL, Python for model training)
  • Brand penetration maps
  • Data engineering (collaboration via Git/Github, CI/CD, ETL, Docker, MLOps)
  • Development of a large-scale end-to-end text classifier (Pytorch, AWS)

Research Assistant

Bocconi Institute for Data Science and Analytics

Aug 2019 – Mar 2020 Milano - Italy
  • Twitter-funded project: Identification of echo chambers and online abuse on the Twitter platform
  • Wordify web-service: Wordify makes it easy to identify words that discriminate categories in textual data. Try it out at wordify.unibocconi.it
  • MACE: a tool for inter-annotators agreement. Try it out at mace.unibocconi.it
  • Topic/Sentiment Analysis
  • Full-Stack web development

Data Science Trainee

European Central Bank (ECB)

Feb 2018 – Nov 2018 Frankfurt am Main - Germany
  • Data Quality Management, Database Management (GUI and API testing), Database Reconciliation
  • Development of Automated Reports and Dashboards (SAP Business Objects, RShiny, Tableau)
  • Data Engineering, Machine Learning, Big Data Processing (Python, R, Spark, Hadoop, Hive, SQL)
  • Selected member of the RIAD–AnaCredit task force with experts from the National Central Banks
  • Head of the RIAD–GLEIF reconciliation project
  • Speaker during regular plenary meetings


MCMC sampling: The Random-Walk Metropolis-Hasting algorithm explained with TensorFlow-Probability



When evaluating redundant annotations (like those from Amazon’s MechanicalTurk), we usually want to (i) aggregate annotations to recover the most likely answer, (ii) find out which annotators are trustworthy, (iii) evaluate item and task difficulty. MACE solves all of these problems, by learning competence estimates for each annotators and computing the most likely answer based on those competences.


Neural Process implementation in Pytorch


Wordify is an online textual processing tool that allows to find out which terms are most indicative for each dependent variable values.

Statistical Concepts with R

Book written with bookdown about elementary statistical concepts.