RehearsAI

What it does

My application is an AI-powered mock interview web application designed to help users practice interviews efficiently. Built using Next.js, the Gemini API, an open-source VAD (Voice Activity Detection), the Deepgram TTS (Text-to-Speech) model, and a touch of LangChain tools, this platform offers a seamless interview experience with a voice latency of just 2-3 seconds.
Users can engage in realistic interview sessions with an AI interviewer, capable of pausing and resuming the interview at any time. The application allows users to personalize their experience by setting their name, job description, and interview duration (up to 10 minutes). As the interview time nears its end (within the last 30 seconds), the AI interviewer provides a wrap-up, offering a comprehensive review of the interview.
Post-interview, users receive detailed feedback, including an overall review and specific comments on different parts of the interview, accompanied by recordings of their responses for reference. This feature-rich application ensures users are well-prepared for real-world interviews, making it an invaluable tool for job seekers.