QA Automation Engineer

We’re looking for a tester who treats quality as engineering, not a checkbox. You’d own testing for one or two client engagements that run in production for years, combining careful test design, exploratory testing and maintainable automation across user journeys and APIs.

What you will do

• Work with product and engineering to understand requirements, user journeys and release risks before a feature is built.
• Write and run test plans for real features before they ship, and explore by hand where automation alone cannot answer a question.
• Design risk-based coverage across APIs, integrations and user interfaces, including negative cases, permissions and accessibility.
• Build automated coverage (system/integration tests) where it earns its keep.
• Keep the suite trustworthy in CI: investigate failures and fix or remove flaky, low-value tests.
• Reproduce and triage production issues; turn each one into a regression test.
• Work directly with the engineers building the feature — no separate QA silo.

What you should bring

Skills and experience

• 5+ years testing web applications, including automated tests for APIs and browser workflows in at least one programming language and an established framework.
• Ability to turn a user story into risk-based test scenarios, not a count of test cases.
• Understanding of test isolation, selectors, waits, data setup and cleanup, and what makes a test flaky.
• Ability to investigate a failure using logs, traces, network responses or database evidence.
• Clear, specific bug reports — steps, expected, actual, evidence.
• A nose for the edge cases other people skip.
• Experience with version control, code review and CI, and comfort working in a shared codebase.
• Practical judgment when using AI to generate tests: check the assertions, coverage and stability before accepting the output.

How you work

• Ownership: you see work through from the first question to production and the days after, own your mistakes and fix what breaks without being asked.
• Dependability: you are on time for calls and pairing sessions, meet the dates you commit to and flag a risk as soon as you see it.
• Quick learning: you get productive in an unfamiliar codebase, language or domain by reading the code and docs and asking focused questions.
• Clear communication: your status updates, pull request descriptions and trade-off explanations are short and easy for non-engineers to follow.
• Pragmatism: you pick the simplest approach that solves the problem and know when to stop polishing.
• Helping others: you make the people around you better through careful code review, pairing and sharing what you learn.

Tech stack

Our applications are built in Ruby on Rails and TypeScript/JavaScript, and you should be comfortable reading code and writing automated tests.

• Test automation: Playwright with TypeScript for browser tests, and API tests against JSON endpoints in code or with Postman.
• CI: running and debugging test suites in GitHub Actions, using traces, screenshots and logs.
• Ruby on Rails: Minitest or RSpec, Capybara system tests, and fixtures or factories. A plus, not a requirement.

Helpful experience

• Shaping a test framework, or making a CI suite faster and more reliable.
• Accessibility, performance, security or mobile testing, with tools such as axe, k6 or OWASP ZAP.
• Improving a product after release using user feedback or production signals.

Working with AI

AI agents write a growing share of our first drafts. The engineering work is in setting them up well, checking what they produce and owning the result. In practice:

• Giving an agent such as Claude Code or Codex the context it needs (AGENTS.md, a written plan, acceptance tests) and having it implement against that.
• Running several agents in parallel on separate branches or worktrees, each on a scoped task, and reviewing each diff before it merges.
• Connecting agents to real tools through MCP, such as the browser, logs, database and issue tracker, so they can check their own work instead of guessing.
• Investigating production issues with AI from stack traces and logs, then confirming the cause with evidence before shipping a fix.

You remain responsible for correctness, security, maintainability and protection of confidential information. Knowing when an agent is confidently wrong matters more to us than how fast it produces code.

Our hiring process

Expect a conversation about test work you have owned, a short live exercise in risk analysis and failure diagnosis, and a practical automation task where AI use is permitted. We will tell you the tool rules before each stage and discuss your decisions and verification steps afterward.

How to apply

Apply through the channel where you found this opening. Share your résumé and a brief description of a bug you caught or a quality improvement you personally owned. We welcome links to public work when available, but public code is not required.