Subprocessor List

Effective Date: September 8, 2026

Last Updated: September 8, 2026

This page identifies third-party providers GleanMark may use to process Customer Content and personal information in operating the Service.

1. About This List

GleanMark uses service providers to operate, secure, support, and improve the Service. A provider is a subprocessor only to the extent it processes personal information or Customer Content on GleanMark's behalf. Use of a provider can depend on the Customer's features, configuration, location, and payment method, so not every provider processes every Customer's information.

Providers are subject to contractual and other safeguards appropriate to their role. Our Privacy Policy describes the categories of information and purposes involved.

2. Current Providers

ProviderProcessing purpose
SupabaseCloud database, authentication, storage, and backend infrastructure
VercelWeb application hosting, delivery, edge infrastructure, and page-performance measurement
RailwayApplication, server, and worker hosting
CloudflareBot and abuse protection for public pages and anonymous requests
UpstashBackground job queuing and coordination for applicable features
StripePayment processing, subscriptions, invoicing, and payment fraud prevention
PostHogProduct analytics, feature measurement, and privacy-masked session replay
GoogleAnalytics, authentication, and AI processing for applicable features
SentryApplication error reporting, diagnostics, and performance monitoring
IntercomCustomer support, in-product messages, and communications
ResendTransactional and product email delivery
SlackAlert and notification delivery to a Customer-connected Slack workspace
OpenAIAI inference and generation for applicable features
AnthropicAI inference and generation for applicable features
xAIAI inference and generation for applicable features
Fireworks AIHosted inference for applicable open-model features
OpenRouterAI inference routing for applicable open-model features
TwelveLabsImage embedding for logo and design similarity search
FastlyDomain availability lookups for applicable clearance features
Merriam-WebsterDictionary definitions used in Office Action drafting

Open-model inference may be served by Fireworks AI directly, or routed through OpenRouter to a host it selects for each request. Where routing is used the downstream host can vary, and routing is restricted to hosts that do not retain or train on submitted prompts. We will update this page if a different named host becomes a direct production subprocessor.

3. Changes and Contact

We may add, replace, or remove providers as the Service changes and will update this list to reflect current production use. Standard online Terms do not promise advance notice of every provider change. A signed order form or data processing addendum may provide additional notice or objection rights.

Questions may be sent to privacy@gleanmark.com.