putting research into practice
Granted research projects, ongoing and past — most are co-funded by the EU, the Italian Ministry, or national agencies. They put my work on AI, NLP, and labour-market intelligence into real-world settings.
Funded by Eurostat & Cedefop · EC-ESTAT/2025/OP/0012
Enhances and maintains the Web Intelligence Hub (WIH) system across all 27+1 EU countries, upgrading it with state-of-the-art AI, ML, and NLP techniques.
PRIN — Italian Ministry for Research · 2023–2025
Develops novel NLP algorithms to extract information from web content (e.g. social media) for sentiment analysis and explainable AI in the financial domain.
European Training Foundation · 2022–2026
Develops and enriches online job vacancy (OJV)-based data systems for a "Big Data for Labour Market Intelligence" initiative in developing and ETF partner countries, within EU external relations policies.
Lanbide — Servicio Vasco de Empleo / Prospektiker · LAN/A-37/2024
Transforms the 2022 prototype into a fully-fledged system that analyses skill demand (ESCO) by occupation, based on job offers from major employment portals in the Basque Country.
CFMT — Centro di Formazione Management del Terziario
Studies job advertisements for service-sector managers in Italy, comparing job demand before and after COVID-19 — including source selection, classification, skill extraction, and visualisation.
Italian Ministry for Research · PNRR · 27 partners (15 universities, 5 institutes, 7 companies)
Within Spoke 1, WP 6, Task 6, defined a methodology for integrating big data into official statistics (smart statistics) and developed a working prototype to monitor ageing-related phenomena through AI and Big Data Analytics.
University of Milano-Bicocca — Grandi Infrastrutture grant
Developed an Italian labour market observatory using Big Data Analytics (on AWS) and AI to process millions of online job ads, extracting occupations, skills, trends, skill gaps, and skill mismatches.
Cedefop · AO/DSL/VKVET-JBRAN/WIH-OJA/002/20
Continuation of the Real-Time LMI framework contract: consolidated the online vacancy analysis tool and integrated it with Eurostat Smart Statistics, extended to Switzerland, Liechtenstein, Norway, and Iceland.
European Commission · H2020-SC6 · GA 101004703
Provided policymakers and the public with insights into the evolving demand for new skills and how to adapt education and training systems, including analysis of past, present, and future labour market trends. Project website.
European Union Agency for Cybersecurity (ENISA)
Enriched the ENISA occupation and skill taxonomy using evidence from online job ads, identifying cybersecurity-related professional profiles and skills in IT, ES, IE, FR, and DE.
Eurofound — European Foundation for the Improvement of Living and Working Conditions
Produced a guideline outlining steps, challenges, and methodologies for implementing Big Data Analytics projects, supported by academic literature and case studies.
Lanbide — Servicio Vasco de Empleo / Prospektiker · LAN/A-46/2022
Designed and developed a web-based prototype that analyses ESCO skill demand by occupation from online job vacancies in the Basque Country, including a user self-assessment tool.
Italian Ministry for Research · PRIN 2017L9HJ25 · partners: CRISP-Unimib, LUISS, Foggia, UniMore, UniGenova
Investigated the impact of AI on consolidated legal categories in labour law, plus tax and corporate law, by processing 5M+ Italian online job vacancies in legal professions using NLP, NLU, and XAI methods.
European Commission — EURES · EaSI-EURES VP/2019/010
Supported Member States and EURES partners in establishing high-quality mapping tables for matching job vacancies and CVs on the European Job Mobility Portal. The resulting ISTAT CP-2011 ↔ ESCO bridge is the official EU mapping.
European Commission · 2019-1-RO01-KA203-063214
Studied digitalization in universities and developed the Digital Escalator, a framework for mapping local digital-skills training offers for specific sectors.
European Commission · 2020-1-UK01-KA204-079009
Developed innovative materials for mid-career skill development; online tools helped Skill Coaches guide users in assessing their skills and matching them to ESCO occupations.
European Commission · 2020-1-ES01-KA226-VET-096242
Expanded digital-skills training for adults through online learning, targeting the social-care and arts sectors affected by COVID-19, with an eLearning toolkit for low-skilled workers.
MIUR · PON F/050276/00/X32 · led by Eustema S.p.A.
Explored data analysis and predictive algorithms for budget forecasting and classification of legal cases and documentation in the legal domain.
Assinform, Assintel, Assinter, AICA · with the patronage of AgID and MIUR
Developed and maintained a monitoring system for the demand for ICT-related professional skills in Italy, applying Big Data Analytics and AI (ML and word embeddings) to millions of online job ads to compute a digital-skill rate measuring the pervasiveness of digital competences across occupations.
Fondazione Giovanni Agnelli · with CRISP-Unimib
Integrated administrative archives (MIUR graduate registry and Ministry of Labour mandatory communications) to build a knowledge base for post-diploma analysis, enabling performance evaluation of Italian technical and vocational secondary schools at two years from graduation. eduscopio.it.
Cedefop EU Agency · framework contract led by University of Milano-Bicocca
Set up the EU system for online vacancy analysis covering all 28 EU Member States and 24 languages.
Regione Lombardia · collaboration agreement under art. 15, L. 241/1990
Integrated heterogeneous structured and unstructured sources (Almalaurea, Open Data, mandatory labour communications, online job ads) into a graph-database knowledge base to support university orientation for graduates and school-leavers.
Cedefop EU Agency · AO/RPA/VKVET-NSOFRO/Real-Time LMI/010/14
Explored the feasibility of using online LMI sources to identify skills needs in Europe in real-time, developing and evaluating a functioning prototype across five EU countries.
ARIFL Agency · Regione Lombardia
Defined a formal algorithm for checking data quality and consistency of weakly-structured administrative data — tens of millions of records from mandatory labour communications (COB) — by expressing the problem as AI planning via model-checking.
ARIFL Agency · Regione Lombardia
Defined and implemented a novel technique (built on UPMurphi) for the automatic synthesis of a Universal Cleanser — a repository of all feasible actions capable of cleansing dirty administrative data.
