Replace code reviews with verified intentAviator Verify
Aviator

Aviator MergeQueue

Keep builds green with a highly customizable merge queue

Deploy more PRs while cutting CI runtimes. Easily merge changes into a massive monorepo and run custom CI validations without breaking builds.

What are merge queues?

From startups to Fortune 500s, the world's most effective engineering teams supercharge their developers with Aviator.

NotionAmplitudeSecureFrameCodaSlackFigmaSquareBoschDoorDashLightspeedNotionAmplitudeSecureFrameCodaSlackFigmaSquareBoschDoorDashLightspeed

The problem

Large dev teams + monorepos = slow CI and broken builds

Lengthy end-to-end tests

Re-running entire test suites on every PR causes delayed, sluggish deploys — even when only a tiny part of the codebase changed.

Elusive flaky tests

Intermittent failures from unrelated PR commits are difficult to debug and silently slow every merge by triggering pointless retries.

Outdated PRs cause conflicts

Merging new changes onto branches that have fallen behind mainline produces build conflicts that block releases.

Incompatible parallel changes

Build failures arise when multiple developers land changes that are individually green but break when combined on main.

The solution

Designed for monorepos at any scale

Aviator is a scalable merge queue platform that works with any build tool — Bazel, NX, Pants, Turborepo, Gradle, Maven. Customize merge workflows tailored to your team's use cases.

Run massively parallel dynamic queues with affected targets

Aviator leverages the independent code paths that exist in a large monorepo to create thousands of parallel queues. Each queue validates an independent set of CI checks and merges concurrently — without sacrificing safety.

  • Disjoint parallel queues on the fly
  • Custom required checks per queue
  • Wildcards for required checks

Predictive batches

Smart, configurable algorithms group compatible PRs into predictive batches so you can merge changes at scale and block any failures from reaching mainline.

  • Smart batching minimises CI runtime
  • Block failures before they reach main
  • Fast-forward merge for a linear Git history

Flaky test resilience

Most large test infrastructures get flaky over time. Aviator has built-in configuration to define a tolerance threshold for intermittent failures, automatically retrying within bounds and quarantining repeat offenders.

  • Tolerance thresholds for retries
  • Automatic retries inside the queue
  • Quarantine policies for repeat offenders

Prioritize critical fixes

Move important fixes ahead in the queue without impacting concurrent merges in the disjoint parallel queues. Pause, resume, or reorder the queue with a click.

  • High-priority merges jump the line
  • Auto-update and merge in one step
  • Pause and resume the queue any time

Integrations

Bring your own CI tool

Aviator MergeQueue is CI-agnostic and plugs into any provider you already use. GitHub Actions, GitLab CI, BuildKite, Jenkins, CircleCI, Argo — combine any of them.

Built for enterprise

Trusted by teams of 1,000+ developers

Aviator ships with the security, reliability, and scalability that large engineering organizations require. On-prem deployments, data lake exports, SOC 2 Type II certified.

  • Real-time support over Slack
  • SOC 2 Type II compliant
  • SAML-based authentication

Background

Why do you need a merge queue?

Teams from 10 to 1,000+ developers eventually hit a deploy bottleneck: how to merge an increasing number of PR changes to mainline while sharing the same codebase. A merge queue automates deploy workflows by prioritising PRs, validating CI, handling semantic conflicts, and merging automatically — ensuring builds stay green even at high velocity.

GitHub · GitLab · Bitbucket · BuildKite · Jenkins · CircleCI · Argo

Stop fighting your merge queue

The Hangar — community of DX leaders

Join us at The Hangar

A vetted community for developer-experience (DX) enthusiasts.