“LAMB: Build and Deploy Safe AI Learning Agents in Your Classroom”
Marc Aller, Politecnical University of Catalonia

AI agents can enrich teaching and learning, but “off-the-shelf” tools like NotebookLM or GPT-based apps often raise serious privacy, security, and ethical concerns—especially in the EU. They also tend to lack alignment with specific learning strategies.
This hands-on session introduces LAMB (Learning Agents Manager and Builder), an open-source project that gives schools and universities full control over their AI in education strategy. With LAMB, educators can
- Create custom learning agents without coding, grounded in their own course materials.
- Deliver agents seamlessly through their LMS or virtual campus (e.g. Moodle) using IMS LTI.
- Choose any large language model provider—or run open-source models on premises.
- Ensure student data remains private and secure.
In this session, you’ll build and deploy your first learning agent step by step, and discover how LAMB can make AI both safe and effective in your teaching practice.
More information:
- The workshop is pending confirmation.
- Computer or tablet required
- Places until full capacity is reached
- Thursday 23, 11:30-12:30, Room 12A
Marc Alier, an Associate Professor at the Politecnical University of Catalonia (UPC), specializes in computer science, information systems, e-learning, IT Ethics and more. With a computer science engineering degree and a Ph.D. in Sustainability, he has contributed significantly to the Moodle community and has authored over 160 academic publications. Alier currently serves as the academic director of the Engineering Education Ph.D. program and teaches multiple courses at UPC’s Barcelona School of Informatics.