Trust
Subprocessors
The service providers Apphelm may use to deliver hosting, email, commerce, integrations and Beam AI investigations.
Effective 22 September 2026 · Version 1.3Current list
| Provider | Purpose | Data / location | Optional? |
|---|---|---|---|
| Cloudflare, Inc. | Workers, D1, R2, Queues, Analytics Engine, Turnstile, Email Service and security. Cloudflare does not run AI models for Apphelm. | Account, configuration, telemetry and service data; Cloudflare's global network (a workspace region is a preference, not a storage guarantee) | No — core infrastructure |
| Stripe entities | Checkout, subscriptions, tax/invoice and merchant-of-record services where eligible | Billing/contact and transaction data; global | Only for paid plans |
| GitHub, Inc. | Customer-authorized release context and issue creation | Integration metadata; customer-selected repositories | Yes |
| Linear Orbit, Inc. | Customer-authorized issue creation | Action/evidence summary chosen by customer | Yes |
| Slack Technologies, LLC | Customer-authorized alerts and digests | Notification content selected by customer | Yes |
| OpenRouter, Inc., and the model host it selects | Answers Beam investigations; currently DeepSeek V4.1 Flash, reached through SansaAI, Sansa Vision's own AI service. This is Beam's only model route: there is no fallback model host | The minimized evidence pack for the question asked; route-dependent and may be outside the EEA | Yes; used only when someone in the workspace asks Beam a question |
Change process
Material new subprocessors will be posted here before they begin processing production customer data. Contracted customers may subscribe through support and object on reasonable data-protection grounds within the notice period. If the concern cannot be resolved, the affected feature may be disabled or the agreement terminated under the DPA.