A Real-Time AI Voice Agent for Interactive Interviews

This system introduces an AI-powered-interview platform that helps authentic job interview experiences through hands-on practice. The system integrates techniques of Natural Language Processing (NLP), Speech to Text conversion and vice versa. In addition to this Large Language Models (LLMs) like Gemini is used to evaluate the overall skills, basically the technical and communication of candidates dynamically. Using this model candidates can experience different versions of adaptive questioning constrained with time. Real-time interaction through browser is achieved through WebRTC and Vapi SDK. Whisper is the Open Source speech recognition model turning speech into text, voice engines making replies that sounds like a human. This framework also provides personalized feedback to the candidates through an assessment module large language models. Options are given to candidates for selecting the level of complexity of questions, clear responses at will, expandable mock interviews towards enrichment of skills.