A Hybrid Intelligence Framework for Personalized Career Guidance via LLMs and Reinforcement Learning

Abstract

Employment guidance for college students often lacks personalization and real-time adaptability in dynamic labor markets. This study presents a hybrid computational intelligence framework that integrates large language models with deep reinforcement learning to support personalized career decision-making. The authors designed a multitask learning architecture to process heterogeneous employment-related data and used reinforcement learning to continuously refine recommendations for resume improvement, interview preparation, and career planning. Experimental evaluation yielded 92.4% matching accuracy, response latency below 1.5 s, and a user satisfaction score of 4.8/5. Further analysis showed a positive association between platform engagement and employment outcomes, with an 85% success rate after sustained use.