LLM Applications: From Core Concepts to Secure Deployment

Cagliari, Sardinia

LLM Applications: From Core Concepts to Secure Deployment is a Blended Intensive Programme (BIP) focused on the design, development, and deployment of secure, reliable, and responsible applications based on Large Language Models (LLMs). Through a combination of online preparatory sessions and an intensive on-site experience, the programme offers both solid theoretical foundations and extensive hands-on practice. Students explore core topics such as LLM architectures, knowledge graphs, machine learning, and AI security, with particular attention to challenges like bias, hallucination, data leakage, and prompt injection. The course emphasizes the integration of LLMs with knowledge graphs to enhance reasoning and data reliability, as well as the implementation of trust and security mechanisms, including validation techniques and model auditing. Ethical and responsible AI practices—covering privacy, fairness, and social impact—are central to the programme. Participants engage in lectures, tutorials, keynote sessions, and collaborative project work, gaining experience with modern tools, open-source frameworks, and APIs. Working in international and interdisciplinary teams, students design, implement, and deploy real-world LLM-based solutions, addressing practical challenges in security, reliability, and scalability. The programme also explores emerging trends in AI deployment and industrial applications, preparing participants for future careers in trustworthy AI systems and research in the field.

Program:

The detailed program is available on the Program page.