Arthur Sarazin

Arthur Sarazin is an Information Systems researcher and Senior IT Consultant at Veltys, specializing in design science, open data, and data product development. Guided by Feynman’s principle, “What I cannot build, I do not understand,” he advocates a prototyping-centered approach, considering construction, experimentation, and explainability as essential levers to understand and develop meaningful data and AI applications.

His work focuses on how organizations, governments, and individuals leverage open data to create new digital services, particularly in urban and geographic contexts. Recently, he has explored ways to break the black-box effect of Generative Artificial Intelligence, which answers questions without revealing its underlying reasoning.

Arthur designs and prototypes tools and methods to enhance the interpretability of Generative AI and to build shared, collective usage of data. He is the creator of the Open Data Canvas, a collaborative framework helping stakeholders structure, understand, and leverage open data. He also developed Le graphe de Borges, a tool that visualizes the reasoning of Generative AI when querying large volumes of data and knowledge.

At Veltys, he applies his expertise in design, data, and technology to consulting projects, connecting research, prototyping, and operational implementation.

“With Pierre Lévy, I am convinced that technological progress — which data and Artificial Intelligence are spearheading today —can only realize its potential when it reshapes how humans and machines interact.”