Fraunhofer IAIS, Dresden Branch of the Institute

Short Profile

Intelligent Systems that Work

The Fraunhofer Institute for Intelligent Analysis and Information Systems IAIS, located in Sankt Augustin near Bonn with sites in Dresden and Heilbronn, is one of the leading institutions for applied research in the fields of artificial intelligence (AI), machine learning and generative AI in Germany and Europe.

With around 350 employees, we develop strategies, technologies and solutions for companies, institutions and organizations along the entire value chain. Our teams focus on the requirements of the finance & legal, healthcare, industry & automotive, media, retail, and public sectors.

Across industries, our solutions range from digital transformation, generative AI models and intelligent process automatization to AI assessment and assurance as well as sustainability and resilience strategies. With our extensive training program, consisting of in-depth data science and AI courses with certification options, we qualify people from different disciplines and levels of knowledge.

We drive innovation forward in collaboration with renowned partners throughout Germany and internationally, both in institutionalized partnerships and in key research policy projects. Learn more about our research network.

Technology Offering Generative Artificial Intelligence

Generative AI is transforming the way we work, create, and make decisions – with enormous potential, but also with new requirements. While it boosts efficiency and innovation, it also raises questions regarding data protection, reliability, and transparency.

We conduct research into the secure use of generative AI in text, audio, and image applications. Together with companies and institutions, we identify useful areas of application, support implementation, and ensure secure integration into existing processes.

We develop and train large language models or adapt them to individual requirements. For example, we actively helped design “Teuken 7B.”

 

AI Language Model “Teuken 7B”

The large language model “Teuken 7B”, developed within the OpenGPT-X research project, was trained from scratch on all 24 official languages of the European Union. The open model comprises seven billion parameters and can be used by researchers and companies for their own AI applications.

 

 

SOOFI (Sovereign Open Source Foundation Models) project

In the SOOFI project, Fraunhofer IAIS is driving forward the development of even larger language models in order to create powerful AI solutions that meet Europe's requirements for reliability, data protection, and user security. Our goal is to strengthen European AI sovereignty by delivering transparent, secure, and user-friendly technologies that offer real added value.