Infrastructure and model providers used to operate Consulting AIOS.

This launch disclosure describes the current operating provider posture and should still be reviewed by counsel before being treated as final negotiated customer terms. The active provider set depends on environment configuration, workspace routing, and whether the customer uses managed or customer-supplied model keys.

Vercel

Application hosting, deployment, serverless functions, cron, and production runtime.

Supabase

Authentication, Postgres database, storage, row-level security, and service-role operations.

Stripe

Checkout, subscription lifecycle, billing portal, invoice/payment metadata, and webhooks where live billing is used.

OpenRouter

Default launch auxiliary model route and optional brokered model route where configured for a customer workspace or platform route.

Google Gemini

Optional AI model provider for selected extraction, visual analysis, advisory, or embedding workflows when configured.

OpenAI and Anthropic

Optional model providers where a customer workspace or platform route is configured to use them.

GitHub

Source control and structured markdown skill library management.

No SOC 2 claim yet.

The launch plan calls for sensible security foundations, not a completed SOC 2 or formal enterprise certification program.

Model routing can change.

Routes may change as quality, cost, and customer requirements are validated. Material provider changes should be disclosed.

BYO keys may reduce platform processing.

Where supported, customer-supplied provider keys can route model usage through the customer-controlled provider account.

Read this with the privacy and retention documents.