Agentic AI as Adaptive Partner in Leadership Education: The Design Architecture of a Live Case Study Platform

The Management and Leadership in Schools (MLS) is the National Institute of Education’s (NIE) milestone school leadership development programme. It is a 17-week residential experience for serving Heads of Department, Subject Heads, and Level Heads. Faculty adopt a variety of pedagogical approaches – including case study. The apparent limitation of cast studies is this: they take months to develop properly. They date quickly. And — this is the part that clearly reveals its limitations — participants discuss them but rarely do anything with them. They analyse. They do not act. For a module on self-management and delegation, that is a real limitation. This article describes the redesigning of case study – specifically designing a live case study platform for MLS3311 that uses Agentic AI to do what a static case never could. The platform is interactive. It adapts to each learner's own professional context. And it gives feedback immediately — not the week after, not in a written comment on a reflection submitted three weeks ago. The design incorporates three components: intelligent tutoring research (including my own earlier work, Ng, 2001), andragogical theory, and contemporary Learning Agent frameworks. The platform integrates five sources of input within a single adaptive environment. At its core is a Learning Agent — comprising a Critic, Learning Element, Performance Element, and Problem Generator — that drives a feedback loop grounded in the learner's own professional situation. The article proposes one central idea: AI agents, embedded in a learning platform, can do what static case studies cannot in terms of facilitation and feedback. AI embedded agent responds and adapts. They challenge the learner through prompts or reflective questions. The human facilitator is not replaced — the facilitator's role actually becomes sharper, not smaller. But the nature of what AI makes possible has genuinely changed. This article is an attempt to describe how.