Subprocessors
Last updated: September 6, 2026
Tuned Tensor Limited uses the subprocessors below to provide and operate Tuned Tensor. Each subprocessor is used only where needed for hosting, authentication, billing, storage, model training, analytics, security, support, or operational reliability.
Local CLI and hosted-service scope
This list covers Tuned Tensor-hosted services and operation of this website. A local-only tt CLI run does not send behaviour specs, datasets, prompts, outputs, or model artifacts to these providers merely because you run it. Providers you independently configure for optional integrations are governed by your account and agreement with them. Managed agent inference uses Tuned Tensor, OpenRouter, and its selected model provider even when your training runs locally.
| Subprocessor | Purpose | Location |
|---|---|---|
| Vercel | Application hosting, deployment, edge infrastructure, and optional analytics. | United States / global infrastructure |
| Supabase | Authentication, database, account data, and application records. | United States / selected cloud regions |
| Amazon Web Services | Managed storage, training infrastructure, logs, and model artifacts. | Selected AWS regions |
| OpenRouter | Routes managed TT agent prompts and tool results to the configured model provider for inference. | Depends on the selected model provider and routing |
| Stripe | Payments, checkout, billing records, fraud prevention, and invoices. | United States / global infrastructure |
| OAuth sign-in when a user chooses Google authentication. | United States / global infrastructure | |
| GitHub | Source control, issue tracking, deployment metadata, and operational collaboration. | United States / global infrastructure |
Updates
We may update this list as the service changes. For questions about subprocessors or data protection terms, contact support@tunedtensor.com.