We are looking for an outstanding intern to contribute to our development of innovative artificial intelligence algorithms, geometric deep learning, and more precisely graph neural networks (GNN) for the analysis of data of the international particle physics experiment ATLAS . ATLAS is located at the Large Hadron Collider (LHC) at CERN. Its aim is to reveal the ultimate properties of matter. In 2028, the rate of collisions provided by the LHC to the ATLAS detector will be increased by an order of magnitude. This will allow the ATLAS experiment to study rare physics phenomena that are at the heart of today’s pressing questions in fundamental physics. Unfortunately the current algorithms will not be able to cope with the complexity and rate of the data recorded. New methods, including the use of AI-based techniques are explored by the ATLAS collaboration. The L2IT team is the world leader, with close collaborators in Berkley (US) and Illinois (US), in the development of a geometric deep learning technique that may solve the problem of the analysis of the data from the innermost part of the ATLAS detector (ITk, a new subsystem for charged particle tracking to be installed in 2028) [2-4]. The intern will be fully immersed in our team and contribute to the ongoing R&D project on GNN at L2IT. His/her work will focus on improving the performance of GNNs through the optimisation of the underlying parameters and architecture of our GNNs. The intern will be confronted with an international working environment. The internship will allow you to obtain strong skills in data analysis and innovative machine learning techniques. Access to high-performance computing facilities is provided.