Legal · Sub-processors

Who else is in the loop.

eri is a small operation, but it sits on a handful of third parties to do things the operator shouldn't be reinventing — payments, email, hosting, auth. Here's every sub-processor currently in use, what data they touch, and where they store it.

Last updated · May 21, 2026

Infrastructure sub-processors

• Stripe

  Privacy →

  Purpose

  Subscription billing and payment processing

  Data

  Name, email, billing address, payment instrument metadata, subscription state

  Region

  United States (with EU storage for European customers)

• Supabase

  Privacy →

  Purpose

  Authentication and Postgres database for account state

  Data

  Email, password hash (or OAuth identifier), account metadata, usage counters

  Region

  Per the project configuration — typically US or EU

• Vercel

  Privacy →

  Purpose

  Web hosting for tryeri.ca and dashboard

  Data

  HTTP request metadata (IP, user-agent, route), edge logs

  Region

  Global edge network

• Resend (or equivalent)

  Privacy →

  Purpose

  Transactional email — sign-in, billing receipts, account events

  Data

  Email address, message metadata

  Region

  United States

• Google

  Privacy →

  Purpose

  OAuth sign-in (only if you choose Sign in with Google)

  Data

  OAuth identifier, email, name as provided by Google

  Region

  Global

Model providers

Model providers are listed separately because they only see data when you actually trigger a model call, and only in the configuration you chose. Calls go from your machine to the provider; the operator does not retain request or response bodies.

• Eri Cloud inference

  Privacy →

  Default cloud agent generation — forwarded without retention

• Anthropic (Claude)

  Privacy →

  Optional BYOK or proxied — used when you connect Anthropic

• OpenAI

  Privacy →

  Optional BYOK — used when you connect an OpenAI key

• Google AI

  Privacy →

  Optional BYOK — used when you connect a Google AI key