InterviewOS - AI-Native First Round Interview Infrastructure

A hackathon-ready MVP that converts first-round interviews into programmable, scalable software infrastructure using AI interviewers.

🚀 Features

• Configurable AI Interviewers: Create interviews by defining role, skills, difficulty, and evaluation criteria
• Scalable Interview Sessions: Multiple candidates can interview simultaneously
• Structured Evaluations: Extract machine-readable signals instead of raw transcripts
• Human Review Dashboard: Visualize scores, evidence, and AI recommendations
• Interview as Code: JSON-based interview schemas for consistency

🏗️ Architecture

┌─────────────────┐    ┌──────────────────┐    ┌─────────────────┐
│   Recruiter     │    │   AI Interview   │    │   Reviewer      │
│   Creates       │───▶│   Engine         │───▶│   Dashboard     │
│   Interview     │    │                  │    │                 │
└─────────────────┘    └──────────────────┘    └─────────────────┘
         │                       │                       │
         ▼                       ▼                       ▼
┌─────────────────┐    ┌──────────────────┐    ┌─────────────────┐
│ Interview       │    │ State Machine    │    │ Structured      │
│ Schema          │    │ + Prompts        │    │ Evaluations     │
└─────────────────┘    └──────────────────┘    └─────────────────┘

🛠️ Tech Stack

• Frontend: Next.js 14, React, TypeScript, TailwindCSS
• Backend: Next.js API Routes, Node.js
• Database: SQLite with Prisma ORM
• AI Layer: Mock LLM interface (easily replaceable with OpenAI/Anthropic)
• UI Components: Custom components with Radix UI primitives

📦 Installation

1. Clone and install dependencies:

cd interview-os
npm install

2. Set up the database:

npm run db:push
npm run db:seed

3. Start the development server:

npm run dev

4. Open your browser: Navigate to http://localhost:3000

🎯 Demo Flow

1. Create an Interview (Recruiter)

• Go to /create
• Define job role (e.g., "Backend Engineer")
• Add skills to evaluate (e.g., "API Design", "Databases")
• Set difficulty level (Junior/Mid/Senior)
• Configure interview style (Friendly/Neutral/Strict)
• Add red flags to watch for
• Click "Create Interview"

2. Share Interview Link

• Copy the generated interview link from the dashboard
• Share with candidates: http://localhost:3000/interview?id={interview-id}

3. Candidate Takes Interview

• Candidate enters name and email
• AI interviewer asks role-specific questions
• Questions adapt based on previous answers
• Progress bar shows completion status
• Structured evaluation generated at the end

4. Review Results (Reviewer)

• Go to /dashboard
• View all candidate evaluations
• See skill-by-skill breakdown with evidence
• Review AI recommendation (Proceed/Borderline/Review)
• Access detailed conversation transcripts

📊 Sample Interview Schema

{
  "role": "Backend Engineer",
  "skills": ["API Design", "Databases", "System Design"],
  "difficulty": "Mid",
  "rubric": {
    "API Design": 30,
    "Databases": 35,
    "System Design": 35
  },
  "redFlags": [
    "No real-world examples",
    "Hand-wavy explanations",
    "Cannot explain trade-offs"
  ],
  "style": "neutral"
}

🤖 AI Interview Engine

The system uses a prompt-based approach with structured JSON outputs:

Question Generation

• Analyzes role requirements and candidate level
• Generates contextual follow-up questions
• Maintains interview flow and progression

Answer Evaluation

• Scores responses on 0-100 scale
• Extracts evidence snippets
• Identifies strengths and weaknesses
• Detects red flags from rubric

Final Assessment

• Aggregates individual question scores
• Provides overall recommendation
• Generates explainable evaluation summary

🔧 Customization

Replace Mock AI with Real LLM

Update src/lib/ai-service.ts:

private async callLLM(prompt: string): Promise<any> {
  const response = await openai.chat.completions.create({
    model: "gpt-4",
    messages: [{ role: "user", content: prompt }],
    response_format: { type: "json_object" }
  });
  
  return JSON.parse(response.choices[0].message.content);
}

Add New Interview Types

1. Create new interview schema in seed data
2. Add role-specific prompt templates
3. Configure skill-specific evaluation criteria

Extend Evaluation Metrics

1. Update InterviewEvaluation type
2. Modify evaluation prompts
3. Add new dashboard visualizations

📁 Project Structure

interview-os/
├── src/
│   ├── app/                    # Next.js app router pages
│   │   ├── api/               # API routes
│   │   ├── create/            # Interview creation page
│   │   ├── interview/         # Candidate interview page
│   │   └── dashboard/         # Reviewer dashboard
│   ├── components/ui/         # Reusable UI components
│   ├── lib/                   # Core business logic
│   │   ├── ai-service.ts      # AI interview engine
│   │   ├── interview-state-machine.ts
│   │   ├── prompts.ts         # AI prompt templates
│   │   └── db.ts              # Database client
│   └── types/                 # TypeScript definitions
├── prisma/                    # Database schema & seeds
└── public/                    # Static assets

🚀 Deployment

Vercel (Recommended)

1. Push to GitHub
2. Connect to Vercel
3. Add environment variables
4. Deploy automatically

Docker

FROM node:18-alpine
WORKDIR /app
COPY package*.json ./
RUN npm install
COPY . .
RUN npm run build
EXPOSE 3000
CMD ["npm", "start"]

🔮 Future Enhancements

• Voice Interviews: Web Speech API integration
• Video Analysis: Facial expression and body language evaluation
• Multi-language Support: Internationalization
• Advanced Analytics: Interview performance metrics
• Integration APIs: ATS and HRIS system connectors
• Real-time Collaboration: Multiple reviewers
• Custom Scoring Models: ML-based evaluation

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

📄 License

MIT License - see LICENSE file for details

🆘 Troubleshooting

Database Issues

# Reset database
rm prisma/dev.db
npm run db:push
npm run db:seed

Port Conflicts

# Use different port
npm run dev -- -p 3001

Build Errors

# Clear Next.js cache
rm -rf .next
npm run build

---

Built for hackathons, designed for production.

Transform your hiring process with AI-native interview infrastructure that scales.