🚀 Mercor IP: AI Domain Expert Interviewer

🔗 Live Demo • 🏗️ Architecture • 🗺️ Roadmap • 🤝 Contributing

An intelligent, real-time mock interview platform that dynamically reads your resume and uses AI to conduct a technical, voice-first interview—tailored strictly to your domain expertise.

---

🌟 Key Features

• 📄 Intelligent Resume Parsing & RAG: Upload your PDF resume. The system extracts core skills, chunks the text, and stores vector embeddings via Supabase. The AI uses local Retrieval-Augmented Generation (RAG) to ask deep, context-aware questions.
• 🎙️ Real-Time Voice Interaction (TTS): Uses the native window SpeechSynthesis and SpeechRecognition APIs to create a completely conversational, screen-free phone-call-like experience.
• 📹 Video & Facial Emotion Analytics: Leverages browser-side neural networks (face-api.js) to track candidate micro-expressions and visual confidence in real-time to compute a comprehensive behavioral scorecard.
• 🔀 Dynamic Tech-Stack Adapters: Configure your Interview Type (e.g., System Design, Behavioral) and Experience Level before starting. The AI interviewer adapts its persona and strictness accordingly.
• ⏱️ Timed Sessions: Interviews are tightly enforced to a realistic 10-minute cap with dynamic warnings.
• 📊 Post-Interview Analytics: Generates an automatic scorecard (Technical Depth, Communication) with fair grading metrics, alongside actionable tips for improvement.

---

🚀 Deployment & Local Setup

1. Prerequisites

• Node.js (v18+)
• A Google AI Studio API Key (for Gemini 2.5)
• A Supabase Project (for Authentication & PostgreSQL Database)

2. Run Locally

git clone https://github.com/tanmay-312/mercorIP.git
cd mercorIP
npm install

Create a .env file in the root directory:

NEXT_PUBLIC_SUPABASE_URL=your_supabase_url
NEXT_PUBLIC_SUPABASE_ANON_KEY=your_supabase_anon_key
GOOGLE_API_KEY=your_gemini_api_key

Start the development server:

npm run dev

Navigate to http://localhost:3000.

3. Docker Deployment

A Dockerfile is included for rapid containerized deployment utilizing the Next.js standalone build output.

docker build -t mercorip .
docker run -p 3000:3000 --env-file .env mercorip

4. Vercel / Netlify Deployment

This Next.js App Router project is perfectly optimized for edge deployments.

1. Push this repository to GitHub.
2. Import the project in Vercel or Netlify.
3. Add your NEXT_PUBLIC_SUPABASE_URL, NEXT_PUBLIC_SUPABASE_ANON_KEY, and GOOGLE_API_KEY to the environment variables settings.
4. Deploy!