Sub- processors

The services that receive personal data from GrapeWire on our behalf, what each one receives, and the switch that turns it off.

Last updated: August 23, 2026

This page describes what the software actually does. Where it states a legal position, that position is ours and has not yet been reviewed by counsel.

RecipientPurposeWhat reaches themControlled by
OpenRouter (and the model vendor it routes to — OpenAI by default)Scoring posts for buying intentThe post author's display name and the full post text, plus your watch name, intent sentence and keywords.The deployment's OpenRouter key. Off entirely when scoring is disabled.
AnthropicScoring fallback; reply draftsSame as above. For reply drafts, also your reply template.The deployment's Anthropic key.
StripePaymentsYour email, workspace name and workspace id. Card details are entered on Stripe's page and never stored by us.The deployment's Stripe key.
Mail relay (our own mail server, hello@grapewire.co)Alert emails, support and account mailAlert emails carry the author name, post text, link, score and reason. Support and account mail carry what you sent us.The deployment's mail configuration.
Hosting providerRuns the application and the databaseEverything GrapeWire stores, and its backups.Deployment.
Your alert destinations — Slack, Discord, Microsoft Teams, Google Chat, email, ntfy, your own webhookDelivering alerts where you asked for themThe whole alert: author name, post text, link, score, reason, watch name.You — each channel you configure.

We update this list before a new recipient receives any personal data, and the date above changes when the list does.

The last row is not a sub-processor we chose: those are the destinations you configured, and once an alert is delivered there it is in a system you control, not us.

Questions about a recipient on this list: hello@grapewire.co.

Read the full Privacy Policy →