An AI-Driven Study on Intelligent Job Matching and Upskilling Recommendations for the Unemployed: A Data-Driven Simulation in the Rwandan Labor Market

This paper introduces Labor-AI: a novel, AI-driven simulation framework for matching unemployed African university graduates with relevant jobs and upskilling pathways using publicly available labor market data. We develop an autonomous pipeline that ingests open job postings (e.g. African job boards, LinkedIn aggregates), CV/resume datasets, and online course repositories (e.g. MOOCs, vocational training catalogs) to model supply-demand dynamics. Using state-of-the-art Natural Language Processing (NLP) and recommender-system methods, our AI identifies skill mismatches and suggests targeted courses or credential programs (e.g. digital skills, technical training) to fill those gaps. The simulation is iterated on Rwandan labor data (2023 LFS) to forecast outcomes for youth unemployment and inform policy. We find that in Rwanda’s context, nearly 21% of youth (16-30) are unemployed, with higher rates among degree holders (≈22.7%). Our framework suggests digital and vocational upskilling can improve match rates substantially; for example, scenarios indicate a ~15% reduction in graduate unemployment with targeted AI-recommended training. In-depth analyses reveal persistent sectoral imbalances (e.g. under-supply in ICT vs. demand in services) and highlight the need for equity-focused AI governance. Ethically, we discuss fairness safeguards to avoid biases (gender, rural/urban, educational background) as noted by experts. Policy implications emphasize building inclusive digital ecosystems (digital literacy, data privacy, AI oversight). Stakeholder-specific recommendations are provided for government, universities, and NGOs.