Generative AI-based Co-Pilot Supporting Citizen in Energy Transition by Leveraging the Benefits of HPC

Short description of the innovation study

The mAiEnergy project will build a generative AI-based co-pilot to support citizens in adopting an active role in a just energy transition by leveraging the benefits of High-Performance Computing (HPC). It enhances digital energy literacy, making concepts such as renewables, efficiency, flexibility markets, and smart grids more accessible. Through personalisation and localisation, it guides users on incentives, grants, and programmes. HPC enables efficient processing and analysis of vast, multi-modal energy datasets and, subsequently, the training of our AI models. By utilising open-source vector databases and implementing a Hybrid Retrieval-Augmented Generation framework for accurate data retrieval, we will develop specialised LLMs tailored to the energy sector, significantly improving the model's accuracy, relevance, and performance.

Organisations involved: 

Technology expert: UBITECH (UBI)

HPC expert: University of Innsbruck (UIBK)

Domain expert: FEN Research (FENR) - The Green Energy Center Europe