Thesis Title:
Hybrid Multi-Agent AI Systems for Dynamic Workforce Prediction
Thesis Abstract:
This project investigates how hybrid Artificial Intelligence (AI) models, combining traditional machine learning with multi-agent reinforcement learning (a type of adaptive AI), can be applied to workforce prediction and skills gap analysis in the semiconductor industry. Current approaches to workforce modelling often rely on static forecasting methods, which struggle to capture the complex, adaptive nature of labour markets. By integrating multi-agent systems, the research explores how AI agents can simulate dynamic interactions between skills supply, technological change, and industry demand.
The case study partner for this research is Arm, a leading UK semiconductor company. Using their workforce data, the model will test the effectiveness of hybrid AI in forecasting emerging skills needs, optimising career pathways, and informing long-term workforce planning. The study also embeds principles of responsible AI, including explainability and human-in-the-loop oversight, to ensure outputs are transparent and actionable for industry stakeholders.
Ultimately, the research aims to contribute both a novel technical framework for workforce analytics and practical insights into the adoption of AI-driven decision support systems in critical technology sectors.
Primary Supervisor:
Dr Francesco Ciriello