Energy-Efficient LLMs for Regional Applications

This project aims to enable energy-efficient and locally deployable large language models (LLMs) that can be operated within the region, supporting digitalization needs without reliance on remote cloud infrastructures. The project focuses on making advanced AI technologies usable for regional companies and public organizations by aligning AI system design with energy efficiency and local infrastructure conditions. In this context, the research focuses on LLM inference, meaning the practical use of already trained models to provide services such as text generation, analysis, or decision support, rather than on the resource-intensive process of training new models from scratch.