VORTEX AI: An AI-Powered Mock Interview Platform Using Large Language Models and Hybrid Multimodal Analysis

This paper presents VORTEX AI, a mock-interview simulation platform that supports interview preparation using large language models (LLMs) and multimodal analysis. Unlike platforms that rely on fixed question banks, VORTEX AI generates personalized questions from a candidate’s resume and target role, and supports both HR and technical interview modes. The system combines verbal response analysis with optional facial, vocal, and posture cues to give structured feedback. The platform is built with React.js, FastAPI, the Gemini API, LangChain, and Firebase. This paper describes the system architecture, the resume-analysis and question-generation methodology, the implementation, and the measured runtime performance of the deployed prototype, along with the engineering challenges encountered and the solutions applied.