On-the-Job Training Management Portal with Sentiment Analysis and Training Plan Content Generation using Large Language Model

The effective management of On-the-Job Training (OJT) programs is essential for bridging the gap between academic learning and workplace readiness. This study introduces a web-based OJT Management Portal designed to address common challenges, including inefficient tracking, misaligned tasks, and limited insights into trainee performance. Key features of the portal include automated attendance tracking, logbook management, and mentor feedback analysis using sentiment analysis to evaluate student progress. Additionally, the platform incorporates AI-driven training plan content generation powered by Llama 3-70B, enabling the creation of structured, role-specific learning paths. Usability testing with stakeholders demonstrated strong user satisfaction, with a mean recommendation score of 4.35/5, highlighting its effectiveness. Recommendations for further enhancements, such as mobile responsiveness and improved user interface consistency, are discussed. This portal represents a scalable, innovative solution for modernizing OJT programs, fostering student development, and improving mentorship processes.