The Artificial Intelligence Lab at Universidad Torcuato Di Tella has secured funding from the Alfred P. Sloan Foundation to research and develop an AI-powered tool aimed at improving how scientists create visualizations. The project, led by Emmanuel Iarussi together with Viviana Siless, aims to facilitate the creation of high-quality visualizations through automated suggestions generated by AI. Emmanuel, the project’s principal investigator, is a full-time professor at the School of Business in the Digital Technology program and the Master in Management + Analytics, where he teaches courses on computer science and information visualization.

The creation of graphics is essential across various disciplines, from sociology to particle physics, as it facilitates the identification of patterns in large datasets and enables effective communication of information. However, designing visualizations that are both clear and visually appealing remains a complex, multidimensional challenge. This project will allow users to upload visualizations to a platform that uses computer vision algorithms to automatically recognize the components and data within the image. Additionally, by leveraging large language models and expert recommendations from specialized literature, the system will evaluate different aspects of the visualization automatically, providing both textual and graphical feedback for iterative improvement. 

The project team is composed of Julian Eisenschlos (Google DeepMind), Leonardo Svarc (Neuroscience Lab, UTDT), Paula Feldman (Artificial Intelligence Lab, UTDT), and Martín Sinnona (University of Buenos Aires), who bring expertise in artificial intelligence, design, and data visualization. The Sloan Foundation is a nonprofit organization dedicated to advancing scientific knowledge, supporting research in areas such as science, technology, engineering, and mathematics. Founded in 1934 by Alfred Pritchard Sloan Jr., its mission is to improve society’s welfare through scientific progress.

[ Versión en español ]