New Instant rollback is live in every plan

Ship on Friday.
Sleep on Saturday.

Meridian watches every deploy, catches the regression in seconds and rolls back before your users ever hit it. One timeline for the whole release — commits, flags, alerts and the people behind them.

Start free — 14 days See how it works

No credit card 4-minute setup SOC 2 Type II

Trusted by engineering teams shipping every day

Why Meridian

Deploys should be boring

Most incidents start with a release. Meridian sits between your pipeline and production and makes the risky part of shipping unremarkable.

Rollback in 8 seconds

Set a threshold once. When error rate or latency breaks it, Meridian reverts to the last healthy build and tells the channel what it did — median time to safe is eight seconds.

Release intelligence

Every deploy is annotated with its commits, feature flags, config changes and reviewers. When a graph moves, you already know which change moved it.

Guardrails, not gates

Progressive rollouts, health checks and approval rules run inside your existing CI. Keep shipping continuously without inventing a change-advisory board.

Context

See the diff that actually mattered

Meridian links every production signal back to the exact lines that shipped. No more bisecting a merge queue at midnight — open the release, read the diff, ship the fix.

  • Commit-level attribution. Every alert carries the commits, authors and reviewers of the release that triggered it.
  • Flag-aware. Toggling a flag counts as a release, so a config change is as traceable as a deploy.
  • Works with your stack. GitHub, GitLab, Vercel, Fly, ECS and plain SSH — one webhook, no agent to babysit.

Visibility

One timeline the whole team reads

Support asks what changed. Product asks when it lands. On-call asks what broke. They all get the same answer from the same page, in language they can act on.

  • Shared release notes. Generated from merged PRs, editable before they go out, published to Slack and email.
  • Environment parity. Compare staging and production side by side and see exactly which build is where.
  • Audit trail included. Every approval, override and rollback is recorded and exportable for compliance reviews.
8sMedian time from alert to automatic rollback
12kDeploys shipped through Meridian every day
41%Fewer release-caused incidents in the first quarter
4.9/5Average rating from 312 verified engineering reviews
“We used to freeze deploys every Friday afternoon. Six weeks after turning Meridian on we stopped — not because we got braver, but because a bad release stops costing us an evening.”
Dana Whitfield VP Engineering, Northbeam

Pricing

Priced per engineer, not per deploy

Ship as often as you like. You are never billed for shipping more.

Solo

€0forever

For side projects, solo founders and open-source repositories.

Get started
  • 1 project, 1 environment
  • Deploy tracking and health checks
  • Manual one-click rollback
  • 7-day release history
Most popular

Team

€24/ engineer / month

For product teams shipping to production several times a day.

Start 14-day trial
  • Unlimited projects and environments
  • Automatic threshold-based rollback
  • Progressive rollouts and flag tracking
  • Slack, Teams and PagerDuty routing
  • 12-month history and audit export

Enterprise

Custom

For regulated teams that need their own tenancy and paperwork.

Talk to us
  • Everything in Team
  • SAML SSO and SCIM provisioning
  • Self-hosted or single-tenant cloud
  • 99.99% SLA and named support

All plans include unlimited deploys · Cancel any time · Annual billing saves 20%

FAQ

Questions engineers ask first

How long does setup actually take?

About four minutes. You connect your Git provider, point one webhook at Meridian and pick the metric that defines “healthy” for your service. There is no agent to install and nothing to run inside your containers.

What happens if Meridian goes down?

Nothing. Meridian sits beside your pipeline rather than inside it, so your deploys continue exactly as they would without us — you simply lose the automatic guardrails until we are back. We publish uptime and every incident review publicly.

Does it work with our existing monitoring?

Yes. Meridian reads health signals from Datadog, Grafana, New Relic, Sentry, CloudWatch or any endpoint that returns a number. Most teams keep their dashboards and use Meridian purely as the decision layer on top.

Can we self-host it?

On the Enterprise plan, yes — as a single container plus Postgres, in your own VPC. Everything else runs in EU or US regions of your choosing, and no source code ever leaves your infrastructure on any plan.

What counts as an engineer for billing?

Anyone who triggers or approves a deploy in a 30-day window. Read-only viewers — support, product, leadership — are free and unlimited, because visibility should not be a paid seat.

Is there a free trial, and what happens after?

Fourteen days of the Team plan with no card required. When it ends you can add a card or drop to the free Solo plan — nothing is deleted and no deploy is ever blocked because a trial expired.

Your next bad deploy is already written

Give it eight seconds to live. Connect a repository and watch Meridian catch the first regression this week.

Start free — 14 days Book a walkthrough