GitBook is a documentation platform for teams building product docs, developer portals, help centers, and internal knowledge bases. We help teams create, maintain, and publish knowledge that is useful for both people and AI-powered workflows.

We chose Upstash Redis because it fits naturally with GitBook’s architecture. We run cloud-based, serverless workloads across both Node.js and edge runtimes, so we needed Upstash Redis to work consistently across those environments without adding operational complexity.

Today, Upstash Redis supports shared backend caching, API rate limiting, and coordination of temporary state across services and workers. We use it to cache expensive work, support reliable rate-limited paths, and keep common backend workflows fast.

The main benefit has been having Upstash Redis available as a managed, serverless primitive that matches how we build and deploy GitBook. It lets our engineering team use Upstash Redis where it is the right tool, while staying focused on product work instead of infrastructure management.

As GitBook scaled cache-heavy workloads, Upstash Redis Enterprise gave the team a better fit for high-traffic usage with predictable economics, no upper bandwidth ceiling, and enterprise-grade features while preserving the serverless model their engineering team already used.