AI Bot Monitor - Vercel Integration Documentation
Overview
AI Bot Monitor is a high-performance middleware solution that tracks AI bot visits to your Vercel-deployed website. Discover exactly what prompts led users to your site with non-blocking logging that maintains optimal performance.
Features
- Real-time tracking of ChatGPT, Perplexity, and other AI bot visits
- Non-blocking middleware architecture (sub-100ms response times)
- Flexible storage options: MongoDB Atlas, Vercel KV, or webhook endpoints
- Intelligent bot detection using advanced user-agent patterns
- Built-in analytics API endpoints
- Automatic MongoDB Atlas configuration through Vercel marketplace
Quick Start
1. Install the Integration
Add AI Bot Monitor to your Vercel project through the marketplace integration.
2. Configure Storage
Choose your preferred storage option:
- MongoDB Atlas: Automatic setup through Vercel marketplace
- Vercel KV: Redis-based storage for high performance
- Webhook: Send data to your custom endpoint
3. Deploy
The middleware automatically starts tracking AI bot visits upon deployment.
Configuration
// middleware.ts
import { createAIBotMonitor } from '@getaiso/ai-bot-monitor';
const monitor = createAIBotMonitor({
storage: {
type: 'mongodb', // or 'kv', 'webhook', 'console'
mongodb: {
database: 'botlogs',
collections: {
requests: 'all_requests',
botVisits: 'visits'
}
}
},
logging: {
logAllRequests: true,
botDetection: true,
excludePaths: ['/api', '/_next']
}
});
export default monitor.middleware;API Endpoints
GET /api/bot-logs
Retrieve AI bot visit analytics
GET /api/all-requests
Get comprehensive request logs
GET /api/analyze-requests
Analyze traffic patterns and trends
Bot Detection Patterns
AI Bot Monitor automatically detects the following AI bots:
- ChatGPT: OAI-SearchBot, ChatGPT-User, GPTBot
- Perplexity: PerplexityBot, Perplexity-User
- Custom Patterns: Add your own detection patterns
Performance
AI Bot Monitor is designed for maximum performance:
- Non-blocking logging prevents middleware delays
- Sub-100ms response time overhead
- Optimized for Vercel Edge Runtime
- Minimal memory footprint
Support
Need help? Check out our support resources: