AI test automation covers two things. First, using AI-assisted and low-code tools to build and maintain automated tests faster. Second, testing the AI features that products now ship — chatbots, summaries, smart search — whose answers vary every time. Both need a tester's judgement more than ever.
"AI tester jobs" gets about 590 searches a month in India and 880 in the US; "qa engineer jobs" 880 / 1,600.
Self-Healing Test Garden
Your tests are little plants in a garden. Press "Redesign the website" and some wilt as locators break. An AI gardener rushes in to heal them — most perk up, but one now grows in the wrong pot (it targets the wrong button). Spot it and move it back. Then a new flower appears: a chatbot feature. Water it with property checks (length, topic, no leaked secrets) instead of exact matches, and it blooms despite changing answers. Pure browser fun, nothing stored.
The test garden
The tools, side by side (checked 23 Oct 2026)
| Katalon | mabl | testRigor | Playwright (+ test agents) | Selenium | |
|---|---|---|---|---|---|
| Style | Record + script | Cloud low-code | Plain-English steps | Code, AI-assisted | Code |
| Free option | Free Studio | 14-day trial; local runs free | Not verified — check site | Free, open source | Free, open source |
| Paid from | Per seat, per year (Professional); see katalon.com/pricing | Quoted | Check site | — | — |
| Best for | Testers moving from manual | Teams wanting cloud runs | Non-coders | Coders and AI-assisted drafting | Existing suites, job ads |
Playwright's documentation describes three test agents — planner, generator and healer — that work with your AI tool of choice.
Slide to your coding appetite
Testing AI features: the new part
When a product has an AI feature, exact-match assertions fail. Instead:
- Check properties: length, required facts, forbidden phrases, no leaked instructions.
- Use repeat runs and report pass rates.
- Keep a small golden set of questions with must-include / must-not-include notes.
- Add safety cases that should be refused.
When this becomes a full-time focus, it's LLM evaluation engineering.
Who takes this route
Manual testers start with low-code tools and their regression lists. Automation testers add AI drafting, healing and AI-feature tests to existing skills.
A sensible first suite
Don't start with the hardest flow. Pick a demo app that invites testing, and automate in this order: login, a search, adding an item, a form with validation errors, and logout. Five flows, four tests each, gives you twenty. Name tests so a stranger understands them ("checkout rejects expired card"). Keep test data separate from steps. Add one deliberately failing test and show how the report explains it. Then add the AI-feature checks last, once the basics run reliably. This order also mirrors how many teams adopt automation, so it makes a natural interview story: what you chose first, why, and what you'd do next.
Reviewing what AI writes
Three habits: (1) read each generated assertion and ask "does this prove the requirement?"; (2) break the feature on a test build and confirm the test fails; (3) review every healed locator. These habits are what separate a tester from a tool operator.
Reading the market
Collect ten job ads near you. Count the tools named, whether AI testing appears, and the experience asked. Let that guide your tool choice.
Pay in test automation
We haven't verified pay data for these roles.
Your realistic 90 days
Manual testers and junior automation engineers come here. Playwright, plus AI helpers that write and repair tests, make automation reachable without years of Java. Your starting artefact is the regression suite you executed by hand.
Assumes 2–3 hours a day alongside job applications, a laptop and a phone, and free plans only. Every step has a smaller version and a way to pause. Use them. It runs in three blocks of 30 days, each ending with a signal that you're on track. A signal is not a promise of a job or income.
What AI changed in your old tasks
| You used to… | Now you… |
|---|---|
| Ran regression by hand each sprint | Run it as Playwright tests on every push |
| Stored cases in Excel | Store them in GitHub with readable names |
| Updated steps after UI changes | Fix flaky locators, with AI suggestions you verify |
Your first 7 days
| Day | Task | Done when |
|---|---|---|
| 1 | Rewrite 8 regression cases you ran often, against a public demo shop (no company data) | 8 cases |
| 2 | Install Node and Playwright; run the sample test | Green run |
| 3 | Record case 1 with codegen; tidy names and assertions | Clean test |
| 4 | Ask a free AI assistant to draft case 2; fix what it got wrong | Test + note of fixes |
| 5 | Add cases 3–4 | 4 passing |
| 6 | Push to GitHub with a README mapping test → manual case | Repo |
| 7 | Save 8 'QA automation', 'Playwright', 'SDET' postings | 8 links |
Days 1–30: Your regression, automated
| Step | Time | Done when | Smaller version | If you need to pause |
|---|---|---|---|---|
| Automate all 8 cases + mapping README | about 14 hours across 3 weeks | 8 green tests | 4 tests. | Leave for later “automate all 8 cases + mapping README” for a fortnight. |
| Ask 4 ex-teammates to review the repo | about 2 hours | 4 messages | 2 people. | Put on ice “ask 4 ex-teammates to review the repo” until next month. |
Signal by day 30: An ex-colleague runs your repo successfully.
Days 31–60: CI and reports
| Step | Time | Done when | Smaller version | If you need to pause |
|---|---|---|---|---|
| Run tests in GitHub Actions with the HTML report published | about 8 hours | CI badge | Manual runs. | It is fine to hold “run tests in GitHub Actions with the HTML report published”. |
| Apply to 10 automation roles listing Playwright or Selenium | about 5 hours | 10 rows | Apply to 5 automation roles listing Playwright or Selenium. | Let go of “apply to 10 automation roles listing Playwright or Selenium”. |
Signal by day 60: A screening call.
Days 61–90: Live-coding ready
| Step | Time | Done when | Smaller version | If you need to pause |
|---|---|---|---|---|
| Practise writing a test from scratch in 20 minutes, twice a week | about 3 hours | 6 timed sessions | 2 sessions. | Let go of “practise writing a test from scratch in 20 minutes, twice a week” till a lighter week. |
| Ask 2 ex-leads for referrals with your repo | about 1 hour | 2 asks | Ask 1 ex-lead for referrals with your repo. | Feel free to postpone “ask 2 ex-leads for referrals with your repo”. |
Signal by day 90: A technical round.
Worries people bring to this route
- Over 40 in automation? Framework design and flaky-test judgement come with years. Show a suite you'd trust in CI.
- Gap since layoff. A public Playwright repo with dated commits answers it.
- No B.Tech. Testing has many BSc and BCA grads; a working suite matters more.
- English in interviews. Walk through your repo on screen. Code carries the conversation.
- Family wants you to 'move to development'. SDET is development. Many titles say engineer.
- Running low on savings. Manual-plus-automation roles hire faster; take one and grow.
Where the openings are, and what to type
Also worth a weekly look: Instahyre, Cutshort, Wellfound, and iimjobs or hirist for senior roles. Upwork and Contra if you want freelance work.
IT services firms, product companies and testing vendors. Naukri carries many 'automation tester' posts that now ask for AI tools; LinkedIn and Instahyre carry product-company ones.
- LinkedIn search string:
("test automation" OR SDET OR Playwright OR Selenium) AND (AI OR LLM OR "self-healing") - Naukri / foundit keywords: automation tester Playwright, SDET, test automation AI, Selenium Java
Job-hunt moves that matter for this route
- Search 'Playwright', 'automation tester', 'SDET' on Naukri and LinkedIn.
- Show the manual→automated mapping; it proves test design.
- ISTQB is optional; a working repo matters more.
Your 30/60/90 quests
XP: learning 20 · proof 60 · outreach 40. Ranks: Wayfinder → Builder → Pathmaker. Progress stays in your browser only.
Course and placement scams
"Guaranteed placement" automation courses with big upfront fees, and job offers asking for deposits. Never share OTPs or UPI PINs. Official domains only. Report: cybercrime.gov.in / 1930 (India), ReportFraud.ftc.gov (US).
AI test automation FAQs
Is Selenium outdated? Still widely named in ads. Do I need a certificate? ISTQB CT-AI v2.0 exists; a working suite is stronger proof.
Take the quiz · From: Manual tester · Automation tester · Related: AI evaluator · Productised QA
Quick quiz: does this path fit?
Answer each, then press Check. Every answer explains itself.
Make your Path Card
A shareable reminder of where you're heading. The day count starts the first time you open this page; quest XP comes from your log above.
Scam check before you apply anywhere
- Real job and gig platforms are free to join. Never pay a "joining" or "training" fee.
- Never share an OTP or UPI PIN — no genuine employer needs them.
- Use official domains only; type the address yourself.
- India: report at cybercrime.gov.in or call 1930.
- US: report at ReportFraud.ftc.gov.
Small steps, every day.
Each quest has a smaller version for hard weeks, and a pause button. Nothing is lost.
Keep exploring
Sources
Checked 23 Oct 2026. We opened every link below ourselves. Search volumes on this site come from OpenSEO keyword research (Oct 2026).
| What we checked | Where |
|---|---|
| Playwright test agents | playwright.dev |
| mabl trial | mabl.com |
| Selenium | selenium.dev |
| Katalon pricing | katalon.com |