iX workshop: evaluating and optimising RAG systems efficiently

A two-day intensive workshop titled “RAG-Systems: Efficient Evaluation and Optimization” has been announced by heise online, offering professionals a chance to deepen their understanding of Retrieval-Augmented Generation (RAG) systems. The workshop focuses on teaching participants how to systematically evaluate existing RAG systems, identify weaknesses, and apply advanced techniques such as semantic chunking, hybrid search, query expansion, reranking, and knowledge graphs to improve accuracy, relevance, and efficiency in real-world applications. The workshop is designed to provide hands-on experience with practical case studies, allowing attendees to work in small teams or pairs under the guidance of two experienced trainers. Participants will engage in realistic use cases and develop independent solutions for common challenges faced in production environments. They will compare different optimization strategies, assess their effectiveness, and learn how to combine them strategically to build robust and high-performing RAG applications. The training is led by Steve Haupt and Lilli Huss from andrena objects, who bring extensive experience in software development and expertise in generative AI. Their combined knowledge allows them to offer insights based on real-world projects, ensuring that the content remains relevant and applicable. The interactive format encourages open discussion, enabling participants to share their own application scenarios and explore diverse solution approaches. This collaborative environment supports immediate experimentation with new methods and direct transfer of learning into professional practice. The workshop is tailored for software developers and AI engineers who already work with RAG systems and wish to enhance the performance, relevance, and reliability of these models. Attendees will gain valuable skills in evaluating and optimizing RAG systems using modern frameworks like RAGAS, which enables data-driven performance analysis. Through structured exercises and expert mentorship, they will acquire the tools necessary to refine their RAG implementations effectively. Participants will have ample opportunity to ask individual questions, exchange ideas, and collaborate on problem-solving within smaller groups. This setup fosters a dynamic learning atmosphere, where theoretical concepts are quickly tested and adapted to specific project requirements. The emphasis on practical application ensures that attendees leave with actionable strategies for improving their RAG systems in actual deployment settings. The workshop aims to equip professionals with the technical know-how and strategic thinking required to navigate the complexities of RAG model optimization. By combining theory with real-world examples, it provides a comprehensive approach to enhancing the capabilities of AI-driven applications. As interest in RAG technology continues to grow, this training offers a timely opportunity for practitioners to stay ahead in the rapidly evolving field of artificial intelligence.