Subprocessors

We use the following subprocessors to provide our services.

• Fly.io
  • Used for our infrastructure
• Cloudflare
  • Used for a reverse proxy
• OpenRouter
  • Used for AI response
  • We never use data for training AI models
  • The provider list
    • OpenAI
    • Anthropic
    • Amazon Bedrock
    • Google Vertex (Global)
    • Google AI Studio
    • Azure
    • Friendli
    • Together
    • Fireworks
    • DeepInfra
    • Inception
  • The model list
    • Anthropic Claude models
    • OpenAI GPT models
    • Google Gemini models
    • Z.ai GLM models
    • Inception Mercury models
• VoyageAI
  • Used for generating embeddings
• Turbopuffer
  • Used for storing embeddings of indexed materials
• Firecrawl
  • Used for web search functionality
• Deepgram
  • Used for the speech transcription
• ElevenLabs
  • Used for the speech generation
• fal.ai
  • Used for AI media generation and speech transcription
• Grafana Labs
  • Used for monitoring / alerting
• Better Stack
  • Used for logging / status check
• Upstash
  • Used for durable workflows
• Neon
  • Used for database (PostgreSQL)
• Redis
  • Used for cache