Aegis Prototype
Adaptive Cybersecurity Education
Overview
Aegis is the core prototype emerging from my doctoral research at King's College London. It explores how adaptive learning systems can be redesigned for cybersecurity education — particularly for populations facing structural disadvantage. The system separates what to teach from how to teach it, modelling cognitive and behavioural signals in real time to personalise both content and delivery. Research is ongoing.
Key Features
Adaptive Content Selection
Dynamically selects and sequences learning material based on the learner's current knowledge state
Cognitive Modelling
Models real-time behavioural and cognitive signals to personalise how content is delivered
Phased Learning Progression
Structured learning phases with gates that adapt pacing to individual progress
Prerequisite Mapping
Content is organised in a dependency graph to ensure learners build on stable foundations
Safety Mechanisms
Built-in detection for learner distress, frustration, and stagnation — with automatic recovery routing
LLM Integration
AI-generated lessons shaped by structured delivery profiles and a clear trust boundary
Architecture
1Dual Adaptation
The system separates knowledge adaptation (what to teach) from delivery adaptation (how to teach it), updating both independently each cycle.
2Real-Time Feedback
Learner interactions generate behavioural signals that feed back into the system, allowing continuous adjustment without manual intervention.
3Equity-Centred Design
The architecture is designed to serve populations facing structural disadvantage in tech education — adapting to cognitive conditions, not just performance metrics.