I design and build automated test suites for web and mobile apps, wire them into your CI/CD pipeline, and hand them off documented and maintainable. One senior engineer, start to finish.
$ npx playwright test --project=chromium --reporter=list Running 48 tests using 6 workers ✓ auth/login.spec.ts › signs in with valid credentials (1.2s) ✓ auth/login.spec.ts › rejects an expired session token (840ms) ✓ checkout/cart.spec.ts › applies promo code before tax (2.1s) ✓ checkout/cart.spec.ts › blocks checkout on empty cart (610ms) ✓ search/filters.spec.ts › keeps filters across reload (1.7s) ✓ api/orders.spec.ts › POST /orders returns 201 with id (320ms) ✓ mobile/nav.spec.ts › menu opens on iPhone 14 viewport (1.9s) ✓ a11y/forms.spec.ts › every input has a label (540ms) … 40 more 48 passed (41.3s) To open last HTML report run: npx playwright show-report $
Most engagements start with one of these. Fixed scope where it makes sense, hourly where it doesn't.
No suite yet, or a pile of manual test cases nobody has time to run? I turn them into a real framework: page objects, fixtures, test data, parallel runs, readable reports.
Flaky, slow, or nobody dares touch it? I audit what you have, cut the flake, speed up the runs, and leave the codebase easier to maintain than I found it.
Tests that run on every pull request, with reports the whole team can read and gates that stop broken builds before they reach production.
Client work is under NDA, so here are two things of my own, running in production. Both have public demos. Click around.
An AI social coach for neurodivergent adults. It gives exact words for phone calls, errands, coworkers and tricky messages. Two coach personas decode what people mean, draft replies, write script cards to read from mid-call, and role-play the conversation first, then score how it went.
A rental finder that pulls live listings around Charlotte, NC, enriches each one with lot shade, tree cover, crime grade and commute distance, then scores and ranks them on a weighted 100-point scale. Claude writes a plain-English verdict for every home, with pros, cons and questions to ask the landlord.
The same four steps whether it's a two-week rescue or a six-month build. You always know what's next.
I read your app, your existing tests and your pipeline, then send a short written plan: what to automate first, what to skip, and what it will cost.
Foundations go in before volume: structure, fixtures, data, reporting. Then tests, in priority order, with a green run on every pull request.
Tests run where your code lives. Failures come with traces, screenshots and a one-line reason a developer can act on.
Docs, a walkthrough, and a maintenance guide. I stay available afterwards, but the goal is that you don't need me.
I'm a test automation engineer in Raleigh, North Carolina. When you hire ETA you get one senior engineer who writes the code, runs the pipeline and answers the email. No bench, no handoffs, no account manager.
I work mostly in Playwright and TypeScript these days, with plenty of Selenium, Java, C# and Appium behind that. I use Claude Code heavily to move faster, and I review everything it produces the way I'd review a junior's pull request.
Between client projects I build and run my own products, which keeps me honest about what shipping and maintaining software actually takes.
Heather · Owner, Elite Test Automation, LLC
I reply within one business day. Email is easiest for me.