Upload any PDF document and get instant AI-powered structured analysis — summaries, key entities, dates, obligations, risk flags, and action items.

Live Demo: https://ai-document-analyzer-ea.vercel.app

• Executive Summary — AI-generated 2-3 sentence overview with document type detection
• Entity Extraction — Automatically identifies key parties and their roles
• Date Detection — Extracts all dates with contextual descriptions
• Financial Analysis — Identifies payments, fees, penalties, taxes, and totals with color coding
• Obligation Mapping — Clear table of who owes what, by when
• Risk Flagging — Highlights concerning terms, unfavorable clauses, and missing information (severity: high/medium/low)
• Key Terms Glossary — Legal and technical jargon explained in plain English
• Action Items — Extracted next steps and required actions
• JSON Export — Download full analysis results for integration or record-keeping
• Real-time Processing — No database, no login required — instant results

• Node.js 20.9.0+ (Next.js 16 requirement)
• Anthropic API key from https://console.anthropic.com/settings/keys

# Clone the repository
git clone git@github-personal:eangao/ai-document-analyzer.git
cd ai-document-analyzer

# Install dependencies
npm install

# Create environment file
cp .env.example .env.local

Create a .env.local file in the project root:

ANTHROPIC_API_KEY=your_anthropic_api_key_here
DEMO_MODE=true  # Enable cost control (5K char limit vs 80K)

Environment Variables Explained:

• ANTHROPIC_API_KEY — Required. Your Anthropic API key from console.anthropic.com
• DEMO_MODE — Optional. Set to true to enable cost protection (reduces API token usage by ~80%)

npm run dev

Open http://localhost:3000 to see the app.

npm run build
npm start

# Run all tests
npm test

# Run tests in watch mode
npm run test:watch

# Run tests with coverage
npm run test:coverage

The repository includes 7 sample PDFs in docs/sample documents/ to test the analyzer:

How to Use:

1. Visit the live demo
2. Upload any sample document from the docs/sample documents/ folder in this repository
3. Observe how the AI extracts entities, dates, obligations, risk flags, and financial items
4. Compare results across different document types

ai-document-analyzer/
├── app/
│   ├── api/
│   │   ├── extract/route.ts       # PDF text extraction endpoint
│   │   └── analyze/route.ts       # Claude API analysis endpoint
│   ├── layout.tsx                 # Root layout with metadata
│   ├── page.tsx                   # Main application page
│   └── globals.css                # Global styles
├── components/
│   ├── FileUpload.tsx             # Drag & drop PDF upload
│   ├── AnalysisLoader.tsx         # Step-by-step loading animation
│   ├── ExportButton.tsx           # JSON export functionality
│   ├── RateLimitAlert.tsx         # Countdown timer for rate-limited users
│   ├── ui/                        # shadcn/ui components (managed by CLI)
│   └── dashboard/
│       ├── AnalysisDashboard.tsx  # Main dashboard layout
│       ├── SummaryCard.tsx        # Document summary + type badge
│       ├── RiskFlags.tsx          # Risk severity alerts
│       ├── EntitiesAndDates.tsx   # Key parties and dates grid
│       ├── FinancialItems.tsx     # Financial breakdown
│       ├── ObligationsTable.tsx   # Obligations table
│       ├── KeyTerms.tsx           # Terms glossary
│       └── ActionItems.tsx        # Action items checklist
├── types/
│   └── analysis.ts                # TypeScript interfaces
├── lib/
│   ├── rate-limit.ts              # Dual-layer rate limiter (3 req/hour per IP + 50 req/day global)
│   ├── error-messages.ts          # User-friendly error message generation
│   ├── error-parser.ts            # Frontend error type discrimination
│   └── demo-mode.ts               # Cost control configuration
├── docs/
│   └── sample documents/          # 7 sample PDFs for testing (contracts, invoices, reports, etc.)
└── next.config.ts                 # Next.js 16 TypeScript config

• Contracts
• Invoices
• Reports
• Legal documents
• Financial statements
• Letters
• General business documents

Risk Severity Color Coding:

• 🔴 High — Critical issues requiring immediate attention
• 🟡 Medium — Notable concerns to review
• 🔵 Low — Minor observations

Financial Item Types:

• 💰 Payment — Regular payments (green)
• 💵 Fee — Service fees (blue)
• ⚠️ Penalty — Late fees, penalties (red)
• 📊 Total — Summary amounts (bold)
• 🏦 Tax — Tax amounts (yellow)
• 🎁 Discount — Discounts applied (purple)

The project includes comprehensive test coverage (667 tests, 93.6% coverage):

• Component Tests — All React components including RateLimitAlert countdown timer
• API Tests — Dual-layer rate limiting, error message generation, DEMO_MODE truncation
• Utility Tests — Error parser (39 tests), demo mode (21 tests), rate limiter (39 tests)
• Integration Tests — Full user workflows, error handling, export functionality
• Accessibility Tests — WCAG 2.1 Level A/AA compliance across all components

Test Distribution:

• Rate limiting: 39 tests (per-IP hourly + global daily limits)
• Error messages: 53 tests (100% coverage on user-facing messages)
• Error parser: 39 tests (type-safe error discrimination)
• RateLimitAlert: 30 tests (countdown timer, styling, accessibility)
• Demo mode: 21 tests (cost control configuration)

Coverage: 93.6% statements | 87% branches | 92.7% functions

This application implements three layers of cost protection to prevent unexpected API charges:

1. Per-IP Hourly Limit: 3 requests per hour per IP address

   • Prevents individual users from exhausting API quota
   • Resets on a rolling 1-hour window
   • Returns friendly error with countdown timer

2. Global Daily Cap: 50 requests per day across all users

   • Hard limit to prevent runaway costs
   • Resets daily at UTC midnight
   • Displays "high traffic" message when reached

3. Demo Mode Toggle: Configurable text truncation

   • DEMO_MODE=true — 5,000 characters (~80% cost reduction)
   • DEMO_MODE=false — 80,000 characters (full analysis)
   • Applied before Claude API call to save tokens

When rate limits are reached, users see:

• Clear error messages explaining what happened and why
• Countdown timers showing exact retry time (MM:SS format)
• Portfolio context acknowledging this is a demo limitation
• Different styling for per-IP (amber) vs global (blue) limits

Example Messages:

• Per-IP: "You've reached your personal limit of 3 documents per hour. Retry available in 47 minutes."
• Global: "This demo is experiencing high traffic. Try again tomorrow morning (UTC)."

• File Validation: PDF-only, 10MB maximum file size
• Input Sanitization: All user inputs validated and sanitized
• No Data Persistence: No database, no stored files — privacy by design
• Type-Safe Errors: Discriminated union types for frontend error handling

With current limits and DEMO_MODE=true:

• Per-IP limit: 3 req/hour = 96% reduction vs unlimited
• Global cap: 50 req/day maximum
• Demo mode: 5K chars = ~80% token reduction per request
• Total: ~$0.45/month maximum (vs potentially unlimited)

1. Click "Deploy" button above or go to vercel.com
2. Import your GitHub repository
3. Add environment variables in Vercel dashboard:
   • ANTHROPIC_API_KEY = your Anthropic API key
   • DEMO_MODE = true (optional, enables cost control)
4. Click Deploy

Your app will be live at your-project.vercel.app

MVP (Current):

• ✅ PDF upload and text extraction
• ✅ Claude AI structured analysis
• ✅ Professional dashboard UI
• ✅ JSON export
• ✅ Comprehensive testing

Future Enhancements:

• 📝 RAG (Retrieval-Augmented Generation) — Upload multiple documents, ask questions across them
• 🔐 User authentication (NextAuth.js)
• 💾 Database integration (Prisma + PostgreSQL)
• 📊 Analysis history and saved documents
• 🌐 Multi-language support
• 💳 SaaS monetization (Stripe integration)
• 📧 Email reports
• 🔄 Batch processing

Contributions are welcome! Feel free to open issues or submit pull requests.

1. Fork the repository
2. Create a feature branch: git checkout -b feature/your-feature
3. Commit your changes: git commit -m "feat: add your feature"
4. Push to the branch: git push origin feature/your-feature
5. Open a pull request

This project is open source and available under the MIT License.

Elmar Angao Full Stack Developer | AI Integration Specialist

Specializing in Next.js, TypeScript, and AI-powered applications using Claude API and modern web technologies.

📧 Contact: elmarcera@gmail.com 💼 Portfolio: [Coming Soon] 🔗 LinkedIn: linkedin.com/in/elmar-angao 🐙 GitHub: github.com/eangao

• Next.js — The React Framework for Production
• Anthropic — Claude AI API
• shadcn/ui — Beautifully designed components
• Vercel — Deployment platform

⭐ If you find this project useful, please consider giving it a star on GitHub!