Mock Interview AI is a browser-based interactive platform that offers a realistic, AI-driven mock interview experience to help users, especially students and early-career professionals, prepare confidently for real-world interviews. The platform supports various interview types, technical, behavioral, and case-based, and lets users select from AI interviewer personas such as tech leads, HR managers, or product managers, each with unique questioning styles and tones. The interaction is entirely voice-based, powered by Google Gemini API for generating dynamic, context-aware responses and Murf AI for natural voice synthesis. Users can verbally respond to questions, just like in a real interview, and the interviewer replies instantly with personalized feedback. One of the standout features is real-time facial expression analysis using face-api.js, which tracks the user’s eye contact, mood, confidence, and engagement level throughout the session. A live mood meter and engagement indicators offer subtle feedback during the conversation, while a comprehensive post-interview report summarizes strengths, areas for improvement, and emotional consistency. The project was built using a modern stack: Next.js 14 App Router, TypeScript, Tailwind CSS, Radix UI, and custom React hooks for managing state and logic. No backend server is required, sessions are stored locally, and APIs are securely accessed using keys stored in environment variables. We designed Mock Interview AI with accessibility and flexibility in mind: it works seamlessly on both desktop and mobile browsers, includes multilingual support, and features intuitive controls with voice waveforms, dark mode, and responsive layouts.