Subprocessors
Effective date: November 4, 2026
Third parties that process customer data so Polylane can run. The default model tier runs on models hosted by Cloudflare, Baseten, Parasail, OpenCode, and OpenRouter. Anthropic and OpenAI receive data only when a workspace selects that provider.
| Name | Purpose | Data |
|---|---|---|
| Cloudflare | Hosting, storage, model inference, email | All service data |
| Amazon Web Services | Receiving events from connected AWS accounts | AWS account events and connection metadata |
| Baseten | Model inference | Prompts, which can include telemetry, code and thread content |
| Parasail | Model inference | Prompts, which can include telemetry, code and thread content |
| OpenCode | Model inference | Prompts, which can include telemetry, code and thread content |
| OpenRouter | Model inference | Prompts, which can include telemetry, code and thread content |
| Anthropic | Model inference, if your workspace selects it | Prompts, which can include telemetry, code and thread content |
| OpenAI | Model inference, if your workspace selects it | Prompts, which can include telemetry, code and thread content |
| Modal | Metric forecasting | Metric time series |
| Typesafe | Video thumbnails | Video frames and a summary of the issue |
| Sign in with Google, fonts | Email and profile; IP address when fonts load | |
| Stripe | Billing | Email and billing details. Stripe holds card details, we never see them |
| Mixpanel | Product analytics | Name, email, usage events |
| Microsoft Clarity | Session recordings in the console | Clicks and page interactions |
| Attio | CRM | Name, email, workspace |
| Buttondown | Newsletter | Email address |
| Pylon | Customer support | Name, email, support messages |
| Slack | Internal notifications | Name, email, feedback |
Integrations you connect — GitHub, Slack, Sentry, Datadog, your cloud accounts — run against your own accounts under your own agreements. They are not subprocessors in that capacity; Amazon Web Services and Slack appear above only for the ways we use them ourselves.
Bring your own model key and inference runs under your provider agreement, not ours.
Questions: privacy@polylane.com.