A Modular LLM-Driven System for Managing Academic Submissions and Editorial Workflow

In this research, an artificial intelligence based academic paper management system is developed. The major aim of the system is to assist conferences and journals in handling the entire process of academic papers from start to finish. The system offers the functionality of uploading papers, assigning users, assigning reviewers, peer review, editorial decisions, and tracking the submissions through an interactive web interface. One of the most highlighted areas in the system is the incorporation of large language models, Google Gemini, in the peer review process. The AutoReview component is used to analyze the language, structure, and originality of the submitted papers, and the Reviewer Matching component uses artificial intelligence to suggest suitable reviewers according to the semantic similarity between the manuscript information and the expertise of the reviewers. The system is developed by utilizing PHP, MySQL, and PDO for the secure back-end, and the front-end is made responsive by utilizing Bulma CSS and jQuery. The developed system is fully functional and is currently being used in a live production environment to manage real-world conference applications.