mAiEnergy: Generative AI-based co-pilot supporting citizen in energy transition by leveraging the benefits of HPC

AI Technology: Generative Language Modeling 

This proposal undertakes mAiEnergy, an initiative to advance the application of Generative AI and HPC in the energy sector. 

By developing a domain-specific AI co-pilot, mAiEnergy aims to enhance citizen engagement in the energy transition by providing context-aware insights on renewable energy, efficiency measures, flexibility markets, and smart grid technologies. 

The project integrates large-scale multimodal datasets and applies specialized LLMs optimized for energy applications. Through HPC acceleration, mAiEnergy will enable faster, more precise energy-related AI models that enhance retrieval-augmented reasoning, ensuring high contextual accuracy and scientific rigor in energy-related decision-making. 

mAiEnergy is a funded innovation study under the Fortissimo Plus Horizon Europe project, selected through the first open call for innovation studies.

UBITECH (UBI), is a leading SME in digital assistants and conversational AI, contributes its experience in deploying AI-driven knowledge applications but lacks dedicated HPC resources. The University of Innsbruck (UIBK) provides foundational expertise in LLMs, HPC optimization, and large-scale AI training optimization, ensuring efficient model performance and computational scaling. FEN Research (FENR), an energy-focused research center, acts as the domain expert, validating AI-generated insights, knowledge retrieval accuracy, and Q/A performance to ensure scientific and practical relevance in energy applications. Together, this SME-university-research center collaboration ensures that mAiEnergy bridges the gap between cutting-edge AI research, computational efficiency, and real-world energy sector applications while relying on EuroHPC resources to make large-scale AI model training and deployment feasible.