Winter School Modules

This Winter’s pedagogical program offers three specific tracks for graduate management and engineering students, and working professionals aiming to master this next generation of AI applications.

Highlights of the session include an introductory seminar on contextual AI, a workshop on building agentic systems integrating human and machine intelligence, workshops exploring leveraging these systems in production, transportation, and communication, a guest conference on enhancing the explainability of AI models, and an evening conference on the UX/UI challenges of collaborative intelligence.

Context is all you Need

This introductory seminar explores the fundamental importance of developing contextual AI models in fields ranging from communications to transportation to manufacturing. Based on the case studies of our Executive White Paper, participants will explore how different dimensions of context are at the heart of decision making, the current limits of Transformer models in capturing context, and strategies for adjusting agentic systems in real-world conditions to help machine and human agents learn together.

Building Agentic Systems

This initial workshop is designed to help the participants work with large language Models and application frameworks in developing AI agents that can sense, reasons, communicate and act. Under the supervision of a Corporate Project Manager, participants will model realistic scenarii for improving operational processes. Particular attention is given to address the business context of the targeted challenges as well as to keep the human-in-the-loop.

From Case Studies to Use Cases

This hands-on workshop will cover the necessary topics to build business-ready applications of contextual AI in the fields of communication, transportation and production. Participants will review the ongoing research projects of Caltech in these areas, model and an agentic system applying in-context learning and positional encoding techniques to address specific challenges, and document realistic use cases moving forward. .

Enhancing the Explainability of AI models

This workshop addresses the critical importance of explainability, transparency and trust in AI, exploring how responsible AI practices can enhance the performance of agentic systems while aligning with the requirements of regulatory frameworks. Yusuf Barman, Project Manager for Digital Advisory and Engineering Projects at ELCA Group, will tackle practical implementation challenges of data quality, the intricacies of prompt templating and the practice of MLOps. Participants will gain hands-on experience applying best practices to AI use cases and insights into the value of AI transparency, which provides visibility into the workings of AI systems, as well as AI explainability, which is a prerequisite to leveraging human intelligence.

Designing the UX/UI of Collaborative Intelligence

Industry experts will address specific UX/IO challenges of agentic systems. Designing systemic features that enhance trust, explainability, and transparency are not just nice-to-have features, they are critical success factors in building contextual AI. Using examples from their current assignments, the facilitators will discuss the challenges of modeling agentic systems to capture, amplify, and communicate the context cues critical to improving personal and organizational decision-making.