Subprocessors
Cloud Posse engages the following subprocessors to assist in providing the Atmos Pro service. This list is updated as our infrastructure evolves. See Section 14.3 of the SaaS Agreement for notice-of-change terms.
Current Subprocessors
| Subprocessor | Purpose | Data Location |
|---|---|---|
| Vercel, Inc. | Application hosting, edge network, blob storage, and AI Gateway | United States for hosting and storage; AI inference varies by provider and routing |
| Neon, Inc. | Managed PostgreSQL database | United States |
| GitHub, Inc. | OAuth authentication, GitHub App webhooks, and repository integration | United States |
| Stripe, Inc. | Payment processing and subscription management (for customers paying via Stripe) | United States |
| Resend | Transactional email delivery | United States |
| Inngest, Inc. | Background job orchestration | United States |
| Linear Orbit, Inc. | Customer support ticket management | United States |
| Sentry (Functional Software, Inc.) | Error tracking and performance monitoring | United States |
| Honeycomb (Hound Technology, Inc.) | Distributed tracing and observability (OpenTelemetry backend) | United States |
| PostHog, Inc. | Product analytics | United States / customer-configurable |
AI processing through Vercel AI Gateway
Atmos Pro uses Vercel AI Gateway for AI chat and infrastructure summaries. Requests are sent to Vercel, which routes them to an inference provider supporting the selected model. Configured models include Anthropic Claude, OpenAI GPT, Google Gemini, and DeepSeek; the inference provider may differ from the model developer depending on gateway routing.
AI requests can include conversation messages, attachments, and infrastructure context retrieved by authorized tools to answer the user's request. The selected inference provider processes this context to generate a response. Model access is through Vercel AI Gateway; the model names above identify the available model families rather than separate direct API integrations.
Atmos Pro's primary application storage is in the United States. AI inference locations and provider retention depend on the applicable gateway configuration, routing, and provider terms. The application's storage region does not guarantee that all AI processing occurs in that region. See Vercel's AI Product Terms, AI Gateway providers, and Vercel's subprocessor disclosures.
AWS Marketplace
Customers who purchase through AWS Marketplace have their billing relationship administered by Amazon Web Services, Inc. under AWS Marketplace terms.
Contact
For questions or objections regarding our subprocessors, please contact legal@cloudposse.com. Customers subject to the SaaS Agreement have a 30-day objection window for new subprocessors under Section 14.3.
Last updated: September 20, 2026