🤖 Fraud Analysis API

Multi-agent fraud analysis system powered by LangGraph

Quick Start

curl -X POST http://localhost:8000/api/v1/analysis \
  -H "X-API-Key: your-api-key" \
  -H "Content-Type: application/json" \
  -d @conversation.json

curl http://localhost:8000/api/v1/analysis/{analysis_id} \
  -H "X-API-Key: your-api-key"

Features

• ✅ Multi-agent analysis pipeline (11 specialized agents)
• ✅ Audio transcription, image description, PDF OCR
• ✅ IMEI enrichment (optional, requires VPN)
• ✅ Biographical data enrichment (optional, requires API key)
• ✅ Async & sync processing modes
• ✅ Structured JSON output with detailed report

Endpoints

• Webhooks
  • POST /api/v1/webhooks/conversation - Receive conversation and trigger analysis
  • GET /api/v1/webhooks/conversation/{id} - Check conversation status
• Cases
  • GET /api/v1/cases - List cases (with filters)
  • GET /api/v1/cases/{id} - Get case details
  • GET /api/v1/cases/{id}/letters-readiness - Check official letters data
• Insights
  • POST /api/v1/insights/compile - Compile insights from all cases
  • GET /api/v1/insights/statistics - Get statistics
  • GET /api/v1/insights/hotlists - List hotlist types
  • GET /api/v1/insights/hotlists/{type} - Get specific hotlist
• Analysis (Legacy)
  • POST /api/v1/analysis - Start async analysis
  • POST /api/v1/analysis/sync - Run sync analysis
  • GET /api/v1/analysis/{id} - Get results
• Health
  • GET /api/v1/health - System health check
  • GET /api/v1/health/agents - Agents health check