Greta Coraglia

Mathematician turned applied AI researcher specialising in Data Science, ML, and Generative AI, and driven by the challenge of making rigorous research into something practical, responsible, and socially useful.

I co-founded MIRAI with industry experts, driven by the conviction that AI compliance should be built on transparency and fairness, not just on ticking boxes. At MIRAI I lead end-to-end R&D across several projects, combining applied research with hands-on development to turn ideas into concrete, working solutions.

---

What I work on

My work sits at the bridge between research and engineering, ensuring theoretical insights become tested, maintainable, and explainable solutions aligned with regulatory requirements.

ML fairness & classification. In-depth assessments of classification algorithms, analysing performance, robustness, and fairness metrics to inform model selection and deployment decisions.

Generative AI & RAG. Designed and evaluated Retrieval-Augmented Generation pipelines, focusing on retrieval quality, answer faithfulness, and end-to-end system performance. Research findings translated directly into architectural and implementation improvements.

---

Talks & Dissemination

A few of my tries at making math and AI for everyone.

• Dalle macchine alle persone: viaggio nel mondo dell’intelligenza artificiale (j/w I. Masnari, D. Posillipo), 2024 (video — Lecco, Segrate)
• Dati sì, ma dati da chi?, Museo Nazionale S&T Leonardo da Vinci, Open night, September 2024
• An open source tool to detect bias in your predictive algorithm (using logic!): BRIOxAlkemy (about, git repo, my slides at AIxIA 2023)
• Matematica da costruire, Archimede, December 2023
• Dissemination corner in The Reasoner, BRIO: From topology to a logic of uncertainty, November 2023 (here)
• Guest post, Identity types in context (j/w Shreya Arya), on the n-Category Café, July 2022 (here)
• Seminario permanente, Un numero finito di idee sulla definizione di infinito, ZEST!, 2020 (recording)

---

Publications

AI, fairness & logic

• Semantics for trust in probabilistic and opaque systems (j/w Francesco A. Genco and Giuseppe Primiero), in progress

• Evaluating AI fairness in credit scoring with the BRIO tool (j/w Genco, Piantadosi, Bagli, Giuffrida, Posillipo, Primiero), arXiv:2406.03292, May 2024

• A bias detecting tool (j/w D’Asaro, Genco, Giannuzzi, Posillipo, Primiero, Quaggio), Proceedings of the 2nd Workshop on Bias, Ethical AI, Explainability and the role of Logic and Logic Programming, published v., January 2024

Category theory & type theory

• Fibrations of algebras (j/w Davide Castelnovo, Fosco Loregian, Ülo Reimaa), submitted for publication, arXiv:2408.16581
  • recording of a talk Fosco gave at the beginning of the project
  • my talk at HoTTEST, 14 March 2024 (slides, recording)
• A 2-dimensional analysis of context comprehension (j/w Jacopo Emmenegger), TAC, published v., October 2024
  • HoTT/UF Workshop, A 2-dimensional analysis of comprehension, 22 April 2023 (slides)
  • ItaCa Fest, Comonads for dependent types, online, 19 May 2022 (slides, recording)
• Categorical models of subtyping (j/w Jacopo Emmenegger), LIPIcs Proceedings of TYPES 2023, published v., December 2023
  • TYPES 2023, Categorical models of subtyping, 15 June 2023 (recording)
• Fuzzy type theory for opinion dynamics (Adjoint School project tutored by Paige North), ongoing
  • ItaCa Workshop, A fuzzy theory of types, Pisa, 21 December 2022 (slides)
• Context, judgement, deduction (j/w Ivan Di Liberti), arXiv:2111.09438, November 2021
  • TallCat, Coding deductive systems into a 2-category, Tallinn, 5 May 2022 — courtesy of the CSM group
  • Logic and higher structures, A 2-categorical representation of deduction, CIRM Marseille, 21 February 2022 (slides)
  • ItaCa Workshop, Context, judgement, deduction, Genova, 20 December 2021 (recording)
• A categorical perspective on Heyting-valued sets, Master’s thesis, awarded AILA’s Best Thesis in Logic prize, July 2019

---

Background

Previously a postdoc with LUCI Lab at the University of Milan, on the BRIO project, and a member of the Scientific Committee of the ItaCa community. PhD from the Mathematics Department at the University of Genova, under the supervision of Pino Rosolini with the logic group.