On this page

Release Pipeline Overview

Learn about Release Pipelines for sophisticated, multi-stage rollout strategies across environments

Release Pipelines roll out feature gate and dynamic config changes in stages across your environments and infrastructure segments. Each phase targets a specific release segment, and you control when a rollout advances with time intervals or manual approvals. Use them to deploy progressively (for example, dev → staging → prod) and monitor each stage before the next begins.

When to use Release Pipelines

Consider using Release Pipelines when you need to:

  • Implement infrastructure-aware deployment strategies beyond what traditional feature flags offer
  • Safely roll out changes in complex, distributed systems with minimal risk
  • Deploy progressively across environments (dev → staging → prod)
  • Target specific infrastructure segments with precision (e.g., prod-us-west → prod-us-east → prod-eu)
  • Control progression between stages with time intervals or manual approvals
  • Monitor each deployment stage before proceeding to the next
  • Enable instant rollbacks if issues arise during any phase

Especially valuable for:

  • Platform engineering teams managing multi-environment, multi-region infrastructure
  • DevOps practitioners implementing progressive delivery strategies
  • SREs responsible for maintaining system reliability across complex deployments
  • Organizations with mission-critical services that can't afford downtime

Key concepts

Get started with release pipelines

Statsig currently supports Release Pipeline only in its Server Core SDKs. Legacy SDKs continue to work but don't get the full features of release pipelines.

Follow these tutorials to begin implementing Release Pipelines:

Current limitations

  • Release pipeline doesn't support experiments
  • Resalt/Disable/Delete/Archive/Launch actions don't trigger a release pipeline
  • Scheduled rollouts don't work with release pipeline

Was this helpful?