Agenta Sub-processors
Last updated: February 4, 2026
If you have questions, contact security@agenta.ai.
note
This list applies to Agenta Cloud (managed SaaS) only. It does not apply to self-hosted deployments.
Current Sub-processors (EU)
| Sub-processor | Purpose | Location |
|---|---|---|
| Amazon Web Services (AWS) | Cloud hosting provider | EU (Frankfurt) |
| SuperTokens | Authentication provider | EU |
| Daytona | Code execution sandbox | EU |
| PostHog | Product analytics | US (soon EU) |
| New Relic | Application monitoring | EU |
| Crisp | Customer support chat | EU (Global) |
| Stripe | Payments & billing | US (Global) |
| Loops | Marketing email provider | US (Global) |
| SendGrid | Transactional email provider | US (Global) |
Current Sub-processors (US)
| Sub-processor | Purpose | Location |
|---|---|---|
| Amazon Web Services (AWS) | Cloud hosting provider | US (Ohio) |
| SuperTokens | Authentication provider | US |
| Daytona | Code execution sandbox | US |
| PostHog | Product analytics | US |
| New Relic | Application monitoring | EU |
| Crisp | Customer support chat | EU (Global) |
| Stripe | Payments & billing | US (Global) |
| Loops | Marketing email provider | US (Global) |
| SendGrid | Transactional email provider | US (Global) |
FAQ
Do you send my data to LLM providers?
Agenta does not process customer data through LLM providers by default.
When you use the Playground or run evaluations:
- You bring your own LLM API keys (OpenAI, Anthropic, etc.).
- Requests are sent directly from your browser/application to the LLM provider.
- Agenta does not store or proxy LLM API calls in typical production use.
For demo/trial purposes, Agenta may provide demo API keys. For production workloads, use your own keys.
Does this list apply to self-hosted deployments?
No. For self-hosted deployments, you control your own infrastructure and any sub-processors you choose to use.