See where the work stalls

Flux reads your GitHub, Linear, and deploy data. Then it tells you which stage your pull requests are sitting in, and how long they have been there.

Example integrations in this demo

Vercel Stripe Supabase PlanetScale Railway Resend

What Flux measures

Each number traces back to a specific event in your repository: a commit, a review, a merge, a deploy.

01  /  Pulse metrics

The four DORA measures

Deploy frequency, lead time for changes, change failure rate, and time to restore — all four derived from your deploy and incident webhooks.

1260
W1 W4 W8 W12

Deploys per week · last 12 weeks · sample org

02  /  Cycle time

  • Coding8.2h
  • Review11.4h
  • QA3.1h
  • Deploy0.9h

Median hours per stage · 23.6h total · sample org

Where a pull request actually spends its time

Coding, review, QA, deploy. Flux records the median hours a change spends in each one. The longest stage is where next week goes, and on most teams it is not the one people name.

03  /  Team health

Review load, per person, per week

Reviews completed per person, counted weekly, alongside focus-time fragmentation and after-hours commit share.

dmr31
apk18
tsl12
jvn7
rho4

Reviews completed, last 7 days · sample org

One quarter, one org

Every figure below is sample data, generated for this demo. A connected workspace shows the same view against your own repositories.

Engineering overview · Q1 2026

Sample data
Deploy frequency
47 / wk
+12% vs Q4
Lead time
2.4h
−18% vs Q4
Change failure
1.8%
−0.4pp vs Q4
Time to restore
14m
+2m vs Q4
60300
Jan Feb Mar

Deploys per week · 13 weeks

Pricing

Starter is free and does not expire. Paid tiers add repositories, history, and views that span the whole team.

Starter

$0 / month

One developer, one repository.

  • 1 repository
  • DORA metrics
  • 7 days of history
  • GitHub integration
Get started

Enterprise

Custom

SSO, audit logs, and a signed SLA.

  • Everything in Pro
  • SAML SSO
  • Unlimited history
  • Custom integrations
  • Dedicated support
Contact sales

Start with one repository

Point Flux at one repo and it backfills 90 days of history in a few minutes. No card, and nothing to install on your machine.

Built with Raven MCP — design intelligence for AI