Your Personal AI-Powered Interview Coach 🚀
Welcome to Candidly! An intelligent desktop application that provides real-time assistance during interview practice sessions. Get instant AI-generated answers, practice with mock interviews, and receive detailed feedback—all while maintaining your interview flow.
- Live Audio Transcription - Capture and transcribe interview questions as you speak
- Screen Intelligence - Detect questions displayed on your screen using advanced OCR
- Instant Answers - Get AI-generated responses in seconds, tailored to your needs
- Mock Interviews - Practice with AI-generated questions across multiple difficulty levels
- 4 Question Types - Behavioral, Technical, Coding & System Design
- Constructive Feedback - Receive detailed analysis of your answers with specific improvement suggestions
- Multi-AI Support - Choose from OpenAI, Google Gemini, or Anthropic Claude
- Answer Modes - Switch between Full Answers and Quick Hints on the fly
- Session History - Automatically save and review all your practice sessions
- Custom Context - Tailor answers to your specific job role and company
- Global Shortcuts - Control the app without losing focus on your interview
- Node.js 18 or higher
- npm (included with Node.js)
- One API key from: OpenAI, Google Gemini, or Anthropic Claude
- Deepgram API key for audio transcription
```bash cd backend
npm install
cp .env.example .env
npm run dev ```
```bash cd frontend
npm install
npm start ```
- Open the Settings window
- Add at least one AI provider API key
- Add your Deepgram API key (for audio)
- Click Start in the overlay
- Begin your interview practice!
Perfect for practicing real interview scenarios:
- Start Listening - Click the "Start" button
- Speak or Display Questions - Ask questions verbally or show them on screen
- Get Instant Answers - AI detects and answers your questions automatically
- Toggle Modes - Switch between Full Answers and Hints anytime
- Copy to Clipboard - One-click copy of answers and code snippets
Practice with AI-generated questions:
- Click "Mock Interview" - Enter practice mode
- Answer Questions - AI asks you questions; you respond
- Navigate Questions - Use "Next Question" or click "End Interview"
- Get Feedback (Optional) - Request detailed analysis of your performance
Make your workflow seamless:
- `Ctrl+Shift+L` - Start/Stop listening
- `Ctrl+Shift+P` - Pause/Resume
- `Ctrl+Shift+M` - Toggle Answer Mode
- `Ctrl+Shift+C` - Copy last answer
Candidly covers all interview question categories:
Focus on soft skills, past experiences, and how you handle situations.
Example: "Tell me about a time when you had to work with a difficult team member."
- 🎓 Best for: Communication, teamwork, problem-solving
- ✅ What you'll practice: STAR method, conflict resolution, leadership
- 💡 Tip: Use specific examples from your work experience
Test knowledge of concepts, architecture, and best practices.
Example: "Explain the difference between REST and GraphQL APIs."
- 🧠 Best for: Deep technical knowledge, design patterns
- ✅ What you'll practice: System concepts, trade-offs, architecture decisions
- 💡 Tip: Explain the "why" behind your knowledge
Solve programming problems with optimal algorithms and clean code.
Example: "Write a function to find the longest substring without repeating characters."
- 💻 Best for: Problem-solving, algorithm design, coding speed
- ✅ What you'll practice: Algorithms, data structures, optimization
- 💡 Tip: Think out loud and explain your approach before coding
Design large-scale systems with scalability and reliability in mind.
Example: "Design a URL shortening service for 100M+ users."
- 🏗️ Best for: Architecture, scalability, trade-offs
- ✅ What you'll practice: System design patterns, databases, caching
- 💡 Tip: Discuss trade-offs and justify your design choices
- Example Mock Interviews - See real examples from all 4 question types
- Mock Interview Guide - Complete feature documentation
Your data stays safe:
- ✅ Encrypted API Keys - AES-256-GCM encryption
- ✅ Local Storage Only - Session history stored on your machine
- ✅ No Recording - Audio transcribed, not stored
- ✅ Screenshots Not Saved - Processed in real-time only
| Component | Requirement |
|---|---|
| OS | Windows, macOS, or Linux |
| Node.js | v18.0.0 or higher |
| RAM | 4GB minimum (8GB recommended) |
| Storage | 500MB for app + dependencies |
| Internet | Required for API calls |
| Microphone | For audio capture |
- Verify Node.js version: `node --version` (need v18+)
- Check port 3000 isn't in use
- Ensure all API keys are in `.env` file
- Verify `DEEPGRAM_API_KEY` is set
- Check microphone permissions
- Test microphone in system settings
- Confirm at least one AI provider key is configured
- Verify the API key isn't expired
- Check you have API quota remaining
MIT License - Free to use for personal and commercial purposes.
Built with amazing open-source technologies:
- Electron - Desktop app framework
- React - UI library
- Socket.IO - Real-time communication
- Deepgram - Speech-to-text
- OpenAI / Google / Anthropic - AI models