Raphaël Tinarrage

Raphaël Tinarrage is a Research Fellow at the Institute of Science and Technology Austria (ISTA), part of the Wagner group. Previously, he held a postdoctoral position at FGV EMAp (Brazil) and was an Assistant Professor at Université Paris-Saclay (2017–2020). His research focuses on Topological Data Analysis (TDA), including theoretical advancements and practical applications in medical imaging, legal document analysis, and equivariant neural networks.

Education: He earned his PhD from Université Paris-Saclay and Inria, advised by Fred Chazal and Marc Glisse (2017–2020). His dissertation, Topological Inference from Measures and Vector Bundles, introduced novel methods for topological data analysis.

Research Interests: Tinarrage explores topological inference (e.g., stability of persistent homology), combinatorial topology (triangulation, Lusternik–Schnirelmann category), and applications like medical image segmentation and legal text analysis. His work bridges theoretical mathematics with computational tools, emphasizing practical implementations.

Key Publications (2021–2025): Recent articles include LieDetect (2025) for Lie group symmetry detection, Empirical Analysis of Brazilian Supreme Court Precedents (2025), and ZigzagNetVis (2025) for dynamic graph visualization. Preprints include Train-Free Segmentation in MRI (2024).

Teaching & Advising: He taught courses on topology and TDA at FGV EMAp and Université Paris-Saclay. He advised multiple students on projects involving legal NLP and TDA applications.

Labs & Tools: His PhD work is implemented in the velour Python package. Code for other projects is available on his GitHub repository.