From NLP to LLMs to Agentic AI — theory, applications and hands-on practice.
This summer school explores the modern AI stack from classic NLP foundations to large language models (LLMs) and agentic AI focusing on how these pieces fit together to build reliable, real-world systems.
Participants will begin by building a solid foundation in core NLP concepts, including text representation, multilingual processing, and evaluation methods. From there, they will explore large language models (LLMs), learning how transformers operate, how to guide their behavior through prompting and fine-tuning, and how to enhance their outputs using Retrieval-Augmented Generation (RAG) for improved accuracy and trustworthiness. Finally, the course expands into agentic AI, where LLMs can plan, leverage tools, retain memory, and collaborate with other agents to accomplish complex, multi-step tasks.
Throughout the summer school, we emphasize responsible and practical Generative AI, addressing core concerns such as fairness, safety, security, privacy, and robustness. Participants will explore engineering and deployment challenges across cloud, edge, and IoT environments, including monitoring, cost-performance trade-offs, and sustainability. Hands-on labs provide practical experience, from building a small RAG pipeline and evaluating LLM behavior to prototyping single- or multi-agent workflows, bridging theory, applications, and practice in NLP, LLMs, and agentic AI.
By the end of the summer school, participants will gain a comprehensive understanding of the modern AI stack — from NLP fundamentals to large language models and agentic AI — and acquire the practical skills to design, implement, and deploy responsible, scalable AI systems using techniques such as prompting, fine-tuning, RAG, and multi-agent orchestration
Our speakers bring expertise from leading AI and technology organizations, including Red Hat, Microsoft, IBM, and more.
Real feedback from our 2026 inaugural edition.
