This workshop will introduce students to the practical use of multi-agent AI systems and show how multiple agents can be effectively combined to solve complex tasks.
The first part will focus on the principles of designing and orchestrating multi-agent workflows, including task decomposition, communication between agents, and assigning specialized roles. The second part will explore how to create custom skills for agents and how these skills can be applied directly in real-world scenarios such as automation, research assistance, content generation, or decision support.
The third part will cover evaluation, debugging, and optimization of agent systems, helping students understand how to improve reliability, efficiency, and output quality in practice.
The workshop will include interactive elements such as live demonstrations, practical examples, and guided exercises where students will design simple agent workflows and experiment with skill-based task execution. The content will be relevant both for beginner students, who will gain an introduction to agent-based AI concepts, and for more advanced students, who will benefit from practical strategies for building more capable and scalable AI systems.