- Events
- •
iSYS shares AI expertise as speaker at 15th CC Partner Symposium
The 15th Competence Centre and Partner Symposium (CCPAF) will be held this year in conjunction with the 3rd Symposium on Machine Learning and Intelligent Systems (MLIS). The symposium offers a discussion forum for practical questions and application-oriented research topics in computer science, business informatics and geoinformatics. It is supported by the Faculty of Computer Science and Mathematics in cooperation with the faculties of Business Administration, Electrical Engineering and Information Technology, Geoinformation, Tourism, Applied Sciences and Mechatronics, as well as Mechanical Engineering, Automotive Engineering and Aeronautical Engineering.
iSYS as a speaker on site
Kimia Ebrahimi and Lukas Sokcevic are AI researchers at iSYS and will be speaking on the topic of ‘RAG in corporate use’.
Kimia specialises in retrieval-augmented generation (RAG), re-ranking and hybrid retrieval strategies. She develops scalable retrieval pipelines and uses vector and keyword searches specifically in a corporate context.
Lukas focuses on large language models (LLMs) and retrieval architectures for efficient context utilisation. He designs modular architectures for AI-supported applications.
- Time: 4:45 p.m. – 5:15 p.m.
- Location: Room R0.007
- Title: RAG in business: How modern retrieval strategies make AI truly smart
- Abstract:
Retrieval-Augmented Generation (RAG) is a key technology for providing large language models (LLMs) such as GPT-4 with up-to-date, internal company information – without the need for time-consuming fine-tuning. The presentation shows why RAG is particularly relevant for companies: It addresses key weaknesses of classic LLMs such as hallucinations, outdated knowledge and missing references.
The content provides a compact and practical overview of how RAG works – from text segmentation (chunking) and the use of vector databases (e.g. Qdrant, pgvector) to embedding and retrieval. The focus is on advanced approaches such as routing, re-ranking and hybrid retrieval, which significantly increase the relevance and precision of the results.
Practical application scenarios illustrate the use of RAG in internal company bots, for example in HR or IT. A typical system architecture is outlined and explained using examples from industry and open source. Challenges such as token limits and data protection aspects – for example, not storing sensitive content in the model – are also discussed.
The aim of the presentation is to provide technical decision-makers and strategists with a well-founded overview of the professional, data protection-compliant and effective use of RAG.
Who is the CCWI?
The Competence Centre for Business Informatics at Munich University of Applied Sciences, or CCWI for short, is a research institute in the field of ICT (information and communication technologies) in distributed applications and, since 2020, has been part of the newly founded Institute for Machine Learning and Intelligent Systems at Munich University of Applied Sciences.
What does the CCWI do?
The aim of the CCWI is to combine university research and teaching with the specific requirements and problems encountered in practice by commercial enterprises through cooperation with partners from the fields of education and teaching as well as industry. This can be achieved, for example, within the framework of research or development projects or through the joint supervision of bachelor’s and master’s theses.
How will the symposium be organised?
Networking through cooperation is the core idea behind the CCWI, which becomes immediately visible and tangible at the annual Partner Day at Munich University of Applied Sciences.
At the CC Partner Symposium, cooperation partners meet at Munich University of Applied Sciences for a joint get-together, present themselves with informative stands, attend accompanying lectures on innovative topics and have ample opportunity for cross-disciplinary exchange of experiences and intensive networking.