AI Mock Interview System: A Web-Based Framework for Structured Interview Practice and Explainable Answer Assessment

AI Mock Interview System is a web-based platform designed to help students and job seekers practise technical and human-resources interviews in an organized way. Interview preparation needs repeated practice, relevant questions, and feedback that explains how an answer can be improved, but candidates do not always have a mentor who can provide this. The current application provides role selection, interview-category configuration, question presentation, session management, and a results interface using frontend web technologies, Java Spring Boot REST services, and MySQL for relevant application data. A clear distinction is made between this existing workflow and an AI assessment capability, which has not yet been verified in the implementation. To go beyond interface integration, the paper proposes a rubric-guided answer-assessment module that evaluates technical correctness, relevance, completeness, and clarity, and returns feedback linked to evidence in the response. The proposed evaluation compares automated scores with independent human ratings and with baseline approaches using error, agreement, and robustness measures. No experimental accuracy or effectiveness result is claimed, because the assessment component, dataset, and tests remain to be completed. Overall, the paper documents the problem, system architecture, implementation considerations, evaluation plan, limitations, and future work.