Skip to content
View jack89-ML's full-sized avatar

Block or report jack89-ML

Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
jack89-ML/README.md

Hi, I'm Jacopo Peracchio 👋

Open Source Analyst · Computational Social Science · Hybrid Threats & Organized Crime

Typing SVG

I build data pipelines and computational models for the social sciences — from electoral data to open-source intelligence — with research-grade standards: reproducibility, audit trails, documentation.

Profile views GitHub followers Python License MIT


🔭 Research Interests

Hybrid threats · Organized crime · International money laundering · Terrorism & armed groups

I work at the intersection of computational social science and security studies:

  • Hybrid threats & disinformation — influence operations, coordinated inauthentic behavior, astroturfing
  • Organized crime & illicit networks — social network analysis of criminal ecosystems, money laundering typologies
  • Terrorism & armed groups — mobilization dynamics, radicalization pathways, network structures
  • Transnational communities — diaspora networks as both risk and resilience factors

🧠 Method: Social Theory × Multi-Agent Systems

I model social phenomena using explicit sociological theory implemented through LLM-powered multi-agent systems — not black-box prediction. The framework draws on Goffman (identity performance), Weber (rationality regimes), Merton (anomie and adaptation), Bourdieu (field, habitus, capital), and social network theory (Granovetter, Burt — weak ties, structural holes, brokerage).

Each agent carries a theoretical role (theory-driven behavioral rules) and the simulation is calibrated against observed data and validated with human expertise — design principles from my research proposal for a multi-agent simulation of influence operations.

🛠️ What I Do

  • Data engineering for social science — extraction, cleaning, normalization and analysis of public datasets (Italian electoral history, demographics, EU open data)
  • Social Network Analysis — academic co-authorship, influence networks, community detection (networkx, Gephi, pyvis)
  • OSINT & text analysis — automated pipelines, compliant web collection, content and sentiment analysis
  • LLM multi-agent systems — agents for research, monitoring and analysis, grounded in sociological models

🧰 Tech Stack

Skills

📈 Activity

GitHub Streak

Note: the GitHub stats cards and trophy widgets (external Vercel services) are temporarily suspended — they will be restored as soon as the services come back online.


📂 Featured Projects

Project What it does Status
TallyHo Italian electoral time series 1946→today, municipality by municipality, from ministerial open data. Python CLI, CSV/JSON/XLSX/Parquet, 90+ tests, CI, Zenodo DOI ⭐ main
Arachne-Scholar Local-first knowledge graph engine for academic literature via SVO dependency parsing. Runs fully offline on your own hardware active
hybrid-kg-paper Experiments on hybrid knowledge graphs (deterministic NLP + LLM) for scientific text. Reproducibility GED 0.0 research

🎓 Education

2023

2026
🎓 Università degli Studi di Torino
Laurea Magistrale in Sociologia e Ricerca Sociale — Sociology
Sep 2023 – Mar 2026

🏅 Grade: 110/110 cum laude
📝 "Economic Networks and Illicit Practices in the Prato District: a Sociological Analysis of the Chinese Presence between Mafia Infiltration and Community Self-Regulation"
🛠️ Organizational skills · Information analysis · +3
2025 🧭 Università degli Studi di Torino — Future Scenarios Lab
Apr 2025 – Jun 2025 · Certificate

Practical lab building frameworks for future-scenario analysis: identifying which factors/events may correlate and show causal links with a phenomenon in its future development.
🛠️ Information analysis · Risk management · +1
2025 🤖 Università degli Studi di Torino — Algorithms & Society Lab
Feb 2025 – Apr 2025 · Certificate
2019

2023
🏛️ Università degli Studi di Firenze
Scienze Politiche — Political Science
Sep 2019 – May 2023

🏅 Grade: 103/110
📝 "New Technologies, Old Inequalities: a Critical Analysis of Digital Human Resources"

💼 Experience

  • MicroRack Systems — founder & operator: data pipelines, research products, multi-agent automation for analysis work

🌍 Languages

Language Level
🇮🇹 Italian Native
🇬🇧 English B2
🇪🇸 Spanish B1
🇨🇳 Chinese A2 (active/passive)

📫 Contact


Research-grade standards: reproducible, documented, auditable.

Pinned Loading

  1. tallyho tallyho Public

    Serie storica elettorale italiana dal 1946: elezionistorico.interno.gov.it (Ministero dell'Interno DAIT) per comune — CSV/JSON/XLSX/Parquet. Open data, CLI, 90+ test, CI, DOI.

    Python