Subprocessors
These are the third parties that process data on our behalf to deliver W3TURN AI Office, and what each one receives. We update this page when a provider changes.
| Provider | What it does | Data involved | Where |
|---|---|---|---|
| Anthropic | AI model processing for agents and Linda | Prompts, customer content, generated outputs | United States |
| OpenAI | AI model processing, where a customer selects an OpenAI model | Prompts, customer content, generated outputs | United States |
| ElevenLabs | Speech for the voice channel and product narration | The text spoken to the caller | United States |
| Twilio | Telephony, when a customer connects a phone number | Call audio, caller number, call metadata | United States |
| Meta Platforms | WhatsApp Business Platform, when a customer enables WhatsApp | Message content, phone numbers, delivery metadata | United States and Ireland |
| Resend | Transactional and notification email | Recipient address, message content | United States |
| Stripe | Payments for subscriptions and usage | Billing contact and payment data | United States |
| Amazon Web Services | Application hosting and container registry | All application data in transit and at rest | Frankfurt, Germany |
| Timescale Cloud | Managed database | All stored data | Frankfurt, Germany |
| Upstash | Cache, rate limiting and event distribution | Transient technical data | European Union |
| Sentry | Error tracking for the Office backend | Error reports: stack trace, error message, office id and route (no request bodies, headers or cookies) | Frankfurt, Germany |
| Vercel | Web application delivery | Request metadata | Global edge network |
Providers marked as only active when a customer enables the relevant channel receive nothing until that channel is switched on for that office.