NextRound 🎯

Get to the next round.
Practice interviews with rubric-based AI feedback and targeted follow-ups.

NextRound is an AI-powered interview simulator designed for job seekers, students, and professionals. It evaluates your interview answers across multiple dimensions, identifies your biggest weakness, and generates personalized follow-up questions to help you improve.

🔗 Live Demo & Project Details • Built with Next.js & Gemini AI

🚀 What is NextRound?

NextRound transforms interview practice with structured AI feedback that goes beyond generic chatbots:

✅ Rubric-Based Scoring - Get detailed feedback across 5 dimensions (Structure, Clarity, Depth, Relevance, Impact)
✅ Weakness Detection - Instantly identify your lowest-scoring area to focus improvement
✅ Answer Improvement - See a better version of your response with rewrite options (Concise, Confident, Technical)
✅ Targeted Practice Loop - Receive customized follow-up questions addressing your weakness
✅ Session Mode - Complete 5-question mock interviews with comprehensive coaching summary
✅ Voice Mode - AI interviewer reads questions and feedback aloud for realistic practice
✅ Session History - Review past interviews and track your progress over time
✅ Low-Effort Filter - Smart detection prevents meaningless answers from wasting API tokens

💡 How It Works

Configure your interview - Choose role (PM, Engineer, Designer), type (Behavioral, Technical, System Design), and difficulty
Answer the question - Type your response (minimum 50 words, or use voice input)
Get AI evaluation - Receive structured feedback with rubric scores (0-10 scale)
Review your weakness - See your biggest area for improvement highlighted
Improve your answer - Rewrite with AI suggestions (Concise, Confident, Technical styles)
Apply & Re-score - See your score improvement in real-time
Practice targeted follow-up - Answer a new question designed to address your weakness
Repeat and improve - Build confidence through deliberate practice

✨ Key Features

🎯 Single Question Mode

Practice one question at a time with instant feedback
Rewrite and re-evaluate answers to see score improvements
Generate targeted follow-up questions based on weaknesses

📝 Session Mode (5-Question Mock Interview)

Complete structured interview sessions with progress tracking
Adaptive question generation based on your performance
Comprehensive coaching summary with score trends and action items
Question counter and progress indicator

🎤 Voice Interviewer Mode

AI reads questions aloud for realistic interview simulation
Hear your feedback summary in natural speech
Auto-play options for hands-free practice
Audio controls: Stop/Resume anytime

🔍 Smart Low-Effort Filter

Detects meaningless answers ("idk", "not sure", empty responses)
Saves API tokens by not evaluating low-quality submissions
Provides coaching tips specific to interview type
Encourages detailed, thoughtful responses

📊 Interactive Answer Improvement

Compare original vs. improved answers side-by-side
Rewrite in different styles: Concise, Confident, Technical
Copy improved answers to clipboard
Apply changes and see score delta

📈 Session History & Analytics

Review all past interview sessions
Track scores across different interview types
Export history for external analysis
Filter by date, role, or interview type

🛠️ Tech Stack

Next.js 16 - React framework with App Router and Turbopack
TypeScript - Type-safe development
Tailwind CSS - Utility-first styling with dark mode support
Framer Motion - Smooth animations and transitions
Google Gemini 2.0 Flash - Advanced AI evaluation with structured outputs
ElevenLabs API - High-quality text-to-speech for voice mode
Zod - Runtime schema validation
localStorage - Client-side session persistence (no backend required)

Architecture Highlights:

Singleton Audio Manager for reliable voice playback
LRU cache for evaluation results (30-min TTL, 50-item capacity)
AbortController for cancellable AI requests
React Context API for global voice mode state
Responsive design with reduced-motion support

⚡ Setup Instructions

Prerequisites

Node.js 18+ and npm
Google Gemini API key (get one free here)

Installation

# 1. Clone the repository
git clone https://github.com/muhibali/NextRound.git
cd NextRound

# 2. Install dependencies
npm install

# 3. Set up environment variables
cp .env.local.example .env.local
# Edit .env.local and add your API keys

# 4. Start development server
npm run dev

Open http://localhost:3000 in your browser and start practicing! 🎉

🔐 Environment Variables

Create a .env.local file in the project root with these variables:

# Required: Your Google Gemini API key
GEMINI_API_KEY=your_api_key_here

# Optional: Preferred Gemini model (default: gemini-2.0-flash-exp)
GEMINI_MODEL=gemini-2.0-flash-exp

# Optional: Alternative provider (OpenRouter)
USE_OPENROUTER=false
OPENROUTER_API_KEY=your_openrouter_key_here

# Optional: Use mock evaluation for testing (bypasses AI, no API key needed)
USE_MOCK_EVAL=false

# Optional: Demo mode for live presentations (ensures app never crashes)
DEMO_MODE=false

# Optional: ElevenLabs Text-to-Speech (for voice mode)
ELEVENLABS_API_KEY=your_elevenlabs_api_key_here
ELEVENLABS_VOICE_ID=EXAVITQu4vr4xnSDxMaL

Environment Variable Details

GEMINI_API_KEY (required) - Your Google Gemini API key for AI evaluation
GEMINI_MODEL (optional) - Model to use (default: gemini-2.0-flash-exp - latest and fastest)
USE_OPENROUTER (optional) - Set to true to use OpenRouter instead of Gemini
OPENROUTER_API_KEY (optional) - Your OpenRouter API key (if using OpenRouter)
USE_MOCK_EVAL (optional) - Set to true for demos without API calls (returns structured fallback data)
DEMO_MODE (optional) - Set to true for live presentations (enables graceful fallbacks, ensures app never crashes)
ELEVENLABS_API_KEY (optional) - Your ElevenLabs API key for text-to-speech audio feedback
ELEVENLABS_VOICE_ID (optional) - Voice ID from ElevenLabs library (default: Sarah - professional female voice)

🎤 ElevenLabs Setup (Optional)

NextRound can read AI feedback aloud using ElevenLabs text-to-speech. This feature is completely optional.

To enable audio feedback:

Sign up for a free account at ElevenLabs
Get your API key from the ElevenLabs dashboard
Browse the voice library and copy your preferred voice ID

Add both values to your .env.local file:

ELEVENLABS_API_KEY=sk_your_api_key_here
ELEVENLABS_VOICE_ID=EXAVITQu4vr4xnSDxMaL

⚠️ Security Critical:

ElevenLabs API keys must NEVER be used in frontend code
Keys are only accessed server-side in /api routes
Never expose process.env.ELEVENLABS_API_KEY to client components
The app will gracefully disable audio if keys are missing

🔒 Security Note

NextRound keeps your API keys secure:

✅ .env.local is git-ignored and never committed to version control
✅ API keys are only used server-side in /api routes (never exposed to client)
✅ No process.env secrets in client components
✅ Zod validation prevents injection attacks
✅ All user data stays local (localStorage only, no backend database)

See .github/docs/SECURITY.md for detailed security guidelines.

🎬 Demo Mode (for Live Presentations)

Ensure your demo never crashes! Set DEMO_MODE=true in your .env.local file before live presentations:

DEMO_MODE=true

What Demo Mode does:

✅ Gemini API fails → Automatically uses mock evaluation (keeps app working)
✅ ElevenLabs TTS fails → Shows gentle error message (app continues without audio)
✅ Network issues → Graceful fallbacks with user-friendly messages
✅ Rate limits → Friendly notifications instead of crashes

Demo Mode messages are designed to be reassuring:

❌ Without: "Text-to-speech service authentication failed. Please check your API key."
✅ With: "Voice mode is temporarily unavailable. The app will continue working without audio."

Recommendation: Enable DEMO_MODE=true for:

Hackathon presentations
Investor demos
Conference talks
Any live audience presentation

Turn it off (DEMO_MODE=false) for normal development and testing.

🎬 2-Minute Demo Script

For hackathon judges or recruiters:

Setup (30 seconds)

Open homepage → Click "Start Practicing"
Select: Product Manager role, Behavioral type, Mid difficulty
Enable Voice Mode for immersive experience (optional)
Click "Start Interview"

Answer Question (45 seconds)

Read the question: "Tell me about a time you had to prioritize features with limited resources"
Type a sample answer (aim for 100+ words)
Click "Evaluate My Answer"

Review Feedback (30 seconds)

Overall Score - See large score with verdict (e.g., "Strong foundation, room to elevate")
Rubric Breakdown - Review 5 animated progress bars showing dimension scores
Biggest Weakness - Note the highlighted weakness (e.g., "Depth")
Strengths & Growth Areas - Scan key insights
Improved Answer - Compare your response to AI-enhanced version

Practice Loop (15 seconds)

Scroll to Targeted Next Question (designed to address your weakness)
Click "Generate Targeted Question" to continue practicing
(Optional) Visit History page to review past sessions

Key talking points:

"Unlike generic AI chat, NextRound uses a structured rubric like real interviewers"
"The targeted follow-up creates a practice loop focused on your weak spots"
"All data stays local - no signup, no backend, instant feedback"

📚 Documentation

Comprehensive documentation is available in the docs/ directory:

Features - Implementation guides for major features
- Audio Manager & Voice Mode
- Low-effort answer filter
- Session mode
- Weakness targeting system
Bug Fixes - Post-mortems and fix documentation
- Answer re-evaluation flow
- Interview type persistence
- Session mode improvements
- Scoring system fixes
Architecture - Design and architecture docs
- Component structure
- Design system
- Animation system
- Scoring algorithms
Project Guides - Core project documentation
- PROJECT_GUIDE.md - Complete project specification
- QUICK_REFERENCE.md - Developer quick reference
- TYPE_SYSTEM.md - TypeScript type definitions
- SECURITY.md - Security best practices

🌟 Future Improvements

NextRound is actively being developed. Potential enhancements include:

🤝 Contributing

Contributions are welcome! Whether you're fixing bugs, adding features, or improving documentation, we'd love your help.

Quick links:

Contributing Guidelines - How to contribute
Project Guide - Full requirements and architecture
Development Checklist - Track implementation progress
Security Guidelines - API key safety rules
Documentation - Feature docs, bug fixes, and architecture

Development Workflow

# Create a feature branch
git checkout -b feature/your-feature-name

# Make your changes and test
npm run dev
npm run build

# Run type checking
npm run type-check

# Submit a pull request
git push origin feature/your-feature-name

Project Structure

nextround/
├── src/
│   ├── app/              # Next.js app router pages
│   ├── components/       # React components
│   ├── lib/              # Core logic, types, utilities
│   └── context/          # React Context providers
├── docs/                 # Documentation
│   ├── features/         # Feature implementation docs
│   ├── fixes/            # Bug fix documentation
│   └── architecture/     # Architecture guides
└── .github/docs/         # Project guides and security

📝 License

MIT License - feel free to use this project for learning, hackathons, or production apps!

See LICENSE for details.

💙 Acknowledgments

Built with ❤️ for the interview prep community. Special thanks to:

Google Gemini for powerful AI evaluation
ElevenLabs for natural text-to-speech
Vercel for seamless deployment
Next.js team for an amazing framework
All contributors and users who provided feedback

🔗 Links

Live Demo & Project Details - Full project showcase
GitHub Repository - Source code and documentation
Report Issues - Bug reports and feature requests

Questions or feedback? Open an issue or reach out to @muhibali

🎯 Get to the next round. Practice with NextRound.

Name		Name	Last commit message	Last commit date
Latest commit History 75 Commits
.github/docs		.github/docs
docs		docs
src		src
.env.local.example		.env.local.example
.gitignore		.gitignore
CONTRIBUTING.md		CONTRIBUTING.md
README.md		README.md
eslint.config.mjs		eslint.config.mjs
next.config.ts		next.config.ts
package-lock.json		package-lock.json
package.json		package.json
postcss.config.mjs		postcss.config.mjs
tailwind.config.ts		tailwind.config.ts
tsconfig.json		tsconfig.json

muhibali/NextRound

Folders and files

Latest commit

History

Repository files navigation