ServicesQA

Find it beforeyour customer does

End-to-end tests · CI gate · release QA

A suite that catches the regression before your customer does, on the paths that actually cost you money.

Best whenEvery release breaks something that used to work.

Covers
Money and data paths
Runs
On every pull request
Reports
Ranked, reproducible
Handover
Suite you own
  • React
  • Next.js
  • Vue.js
  • TypeScript
  • Node.js
  • Python
  • Astro
  • Bun
  • React Native
  • Expo
  • GraphQL
  • REST
  • Playwright
  • Vitest
  • Figma
  • Burp Suite

01How it runs

Four steps

We test the flows that cost you something when they break, then wire the suite into your pipeline so a regression stops a merge instead of reaching production.

  1. 01

    Find what actually matters

    Checkout, sign-up, permissions, anything that writes to a ledger. Coverage percentage is not the goal — the paths a failure would be expensive on are.

  2. 02

    Automate the critical paths

    Playwright for end-to-end, Vitest underneath. Written to survive a redesign: tests target roles and labels, not the CSS classes of the week.

  3. 03

    Gate the pipeline

    The suite runs on every pull request and a red run blocks the merge. A test suite nobody runs is a file, not a safety net.

  4. 04

    Manual pass on each release

    A human works the release candidate before it ships. Automation catches regressions; a person catches the thing that is technically working and obviously wrong.

02In scope

What lands in your accounts

The suite is yours, in your repo, in the language your team already reads. No test-runner licence and nothing to renew.

  • End-to-end tests on the flows that move money or lose data
  • The suite wired into your pipeline, so a red build blocks the merge
  • A manual pass on each release candidate, not only the robots
  • A written defect report ranked by impact, with steps to reproduce

PricingPriced from the written scope — one figure, agreed before code. No hourly billing, no change-request desk.

03Coverage

What we test, and what we do not

A coverage percentage tells you how much of the code ran, not whether the thing that matters still works. This is the honest version of that number — the list is agreed with you before the first test is written.

We test

  • Checkout, payment and anything that writes to a ledger
  • Sign-up, sign-in, password reset and session expiry
  • Permissions — that a user cannot reach another account
  • Writes that cannot be undone, and the confirmations in front of them
  • Forms long enough that losing them costs a customer
  • The paths your last three incidents came through

We do not

  • Pixel positions — a visual diff fails on every redesign
  • Third-party uptime, which is their status page, not your suite
  • Load and performance, which is measurement, not assertion
  • Whether a screen is well designed, which is a review, not a test

Next step

Tell us what
you need built

Send the idea in whatever shape it is in — a document, a sketch, or two sentences. You get a written scope and a fixed figure back before anything is committed to.