Senior Data Scientist and Econometrician

Progetto Lagrange scholarship. Moncalieri, Italy. In this challenging position I complemented mainstream econometrics and simulation-based approaches into large scale analysis of economic administrative data. The diversity of sources required integrating and harmonizing data from legacy and current systems. The resulting published works present novel and compelling evidence of bounded rationality in economic behaviour.

based at the Harvard Medical School, with Alex Lancaster.

based at the ENS Lyon, with Jean-Pascal Bassino; finding evidence of the unbreakable ties and the nature of "Big-6" keiretsu-style industries in Japan.
CS skills I develop novel algorithms and metrics (graph-based semi-supervised learning through enhanced diffusion regimes, and bridging measures of network centralities are examples of the former and the latter, respectively) from the ground up, and also am a proficient user of the standard Python statistics / machine learning libraries. In developing network-based algorithms, metrics and visualizations, I rely on Python, Java, C, C++, Objective C, Fortran; (non-)relational database solutions (e.g. Postgre/MySQL, MongoDB); my toolkit includes R, Gephi, the Swarm, Repast, Mason libraries; I prefer working under UNIX major flavours, GNU/Linux in particular. In coding Agent-Based Models, I tend to start prototyping in NetLogo, then move to a middle-ground object-oriented language (Python being my current choice), if needed also implementing distributed computing (e.g. Spark), eventually down to a lower-level language (C++). I can also take care of calibration and validation, and ETL. I also had experienced working with the SAS suite (Base/STAT) on legacy data projects.

2016-present "Devising Null Models of Lexicostatistics and Glottochronology", based at the Santa Fe Institute, with David Wolpert, William Croft, Richard Blythe, et al.; Agent-Based modelling of language evolution.

based at the ENS Lyon, with Patrick Flandrin; graphbased reconstruction of the history of science, with a scientometrics approach.

Previous positions

(since 2015), Swarm Development Group.
Teaching Activities

Research position in which - owing to my multidisciplinary experience - I liaised between the business-oriented and the fundamental research computer science teams. My research topics included big data analysis, including mining data for socio-economic modelling and natural language processing, also involving large-scale distributed computing efforts.

Italy.

In this position I multitasked in multiple capacities, spanning from data wrangling/analytics to knowledge transfer, to pure research. I brought novel computational approaches into my field of research, leaving behind standard econometric approaches in favor of algorithmic simulations (far-from-equilibrium dynamics in economics) and graph theoretical metrics, disseminated into a wide body of publications.

Major Ongoing Collaborations 2017-present "Graph-based Semi-Supervised Learning", originally based at the ENS Lyon, with Sarah De Nigris; graph partitioning by biased random walks.

Regional emergency medical services were in need of an overhaul, with shrinking budgets and strict requirements to rationalize expenditure; on this assignment, I developed a model of (heli)ambulance services, providing decision-makers with a tool to experiment with alternate scenarios.

Saarbrücken, Germany; Athens, Greece. As a freelance consultant, in a leading capacity I provided industrial partners with a production-grade supply chain optimization solution. Optimized production plans were computed by a combination of simulated processes evaluation and evolutionary algorithms (Agent-Based models and Genetic Algorithms).
Visiting and other positions

I provided guidance in the project modelling effort, including Agent-Based formalization, calibration and validation, based on data and empirical evidence provided.