Legal
Sub-processors
Effective [EFFECTIVE DATE] · Artemis Design Labs
Draft — pending legal review. This document is being finalized and reviewed by counsel. Bracketed [PLACEHOLDER] values will be completed before launch.
Artemis Design Labs uses the providers below to run Bakufu. On Free and Pro plans, your code is processed on our own infrastructure. It is not sent to third-party AI providers, except during the short failover our Privacy Policy describes. Cloud AI inference otherwise applies only to the paid Experttier, and the product labels each cloud-backed agent. We never use your content to train models.
| Sub-processor | Purpose | Data | Applies to |
|---|---|---|---|
| Google (Google Cloud / Identity) | Sign-in (OAuth) identity | Name, email, profile image | All accounts |
| Cloudflare | Encrypted network transport: the HTTPS entry point that reaches our servers | Connection metadata only, such as IP address and TLS handshake | All accounts |
| Tailscale | Backup network transport, used if the main entry point is down | Connection metadata only, such as IP address and TLS handshake | All accounts (fallback route) |
| Microsoft Azure / Amazon Web Services | Cloud AI inference for Expert agents, and backup inference if our own hardware is down | Account data, and your content — processed, never used to train | Expert tier · backup: only while our own hardware is down |
| Anthropic | Frontier model inference for Expert agents | The content you send an Expert agent so it can produce a result | Paid Expert tier only |
| Stripe | Payment processing | Billing contact and payment details (held by Stripe) | Paid plans (when billing is live) |
We update this list as our providers change. For questions, email privacy@bakufu.ai.