IntervueAI is a full-stack AI-powered interview simulation platform designed to help users improve their communication, confidence, and interview performance. Users can choose interview personas, answer real questions, receive AI-based feedback, and track progress over time.
- Real-time audio recording & transcription
- AI-generated interview questions (persona + difficulty)
- Speech-to-text using Whisper
- AI analysis for communication & confidence
- Dashboard with aggregated insights
- PDF report generation
- Session history & performance tracking
- JWT authentication & secure login
- Responsive UI
- React 18 + TypeScript
- Vite
- Tailwind CSS + shadcn/ui
- React Router v6
- Axios
- jsPDF
- Node.js + Express
- MongoDB + Mongoose
- JWT (Auth)
- bcryptjs
- express-validator
- FastAPI
- Whisper (Speech-to-text)
- Llama3 / Ollama (LLM feedback & analysis)
- requests library
- MongoDB Atlas
- User logs in / signs up
- Chooses persona + difficulty
- AI interview simulation begins
- Audio recorded + transcribed using Whisper
- LLM analyzes answers for clarity, relevance & communication
- Scores & feedback shown on dashboard
- PDF report generated
- Average confidence & communication tracked over sessions
git clone https://github.com/ananyajaiswall/IntervueAI
cd IntervueAIcd backend
npm installCreate .env file:
PORT=5000
MONGO_URI=<your mongodb atlas uri>
JWT_SECRET=<your secret>
Run:
npm startcd ../python_api
python -m venv .venv
source .venv/bin/activate # (Mac/Linux)
.\.venv\Scripts\activate # (Windows)
pip install -r requirements.txtRun:
python interview_api.pycd ../frontend
npm install
npm run dev- Body language detection (CV + Vision Transformers)
- Real-time corrective suggestions
- HR/Tech persona marketplace
- Job role-specific interview templates
- Adaptive difficulty
Designed as an interview preparation tool for:
- Students & freshers
- Job switchers
- Mock interview practice
- Communication improvement