This internship focuses on the AI-assisted design of nanoporous carbon electrodes for next-generation supercapacitors, a key challenge in electrochemical energy storage. Conventional simulations often treat electrodes as rigid, yet recent studies show that electrode flexibility strongly affects charging kinetics and ion dynamics. This flexibility should also modify the local electronic response of the carbon surface, which current constant-potential models only approximate. The project aims to implement a machine-learning description of the electron density, enabling large-scale molecular dynamics simulations of realistic disordered carbons with near–ab initio accuracy. By exploring how pore size distribution, sp²/sp³ ratio, and mechanical compliance influence performance, the student will identify structural features that optimize both energy and power density. This internship is funded by the PEPR DIADEM as part of the DIADEM Academy, and will be carried out in collaboration with NTU Singapore and CNRS@CREATE, ensuring strong international and national visibility. It is designed as the first step toward a PhD thesis, extending the approach to other electrode materials.