SDET notes
11 of 428 written

Interview questions about testing, answered one at a time.

I am preparing for an SDET-2 interview against a syllabus I built from a hundred and sixteen real job descriptions. It is fixed at 428 items across 13 topics, and I do not add to it mid-lesson.

A concept goes up here only after I have explained it back in my own words. Everything below is the map; 11 of them are written so far, and the rest are what I have not reached yet.

Topic 1Quality Engineeringin progress
11 written · 51 items

Crisp definitions + live test-design + metric literacy + own the CI quality gate & evidence-backed go/no-go. STOP before ISTQB cert, TMMi, company-wide QE org strategy.

  • MUST 15
  • SHOULD 30
  • STRETCH 2
  • SKIP 4
  • 6 lessons

Design/own/defend a reliable CI-friendly Playwright framework; Selenium literacy (52% of JDs) + BDD/Cucumber. STOP before self-healing engines, custom runners, grid infra.

  • MUST 17
  • SHOULD 13
  • STRETCH 2
  • SKIP 2
  • 15 lessons
Topic 3API / Backend Testingnot started
33 items

Own REST/auth/JWT/schema/chaining + one strategy layer (contract+mocking+microservices pyramid). #1 cluster (69%). STOP before OAuth crypto internals, SOAP mastery, Kafka internals.

  • MUST 14
  • SHOULD 10
  • STRETCH 7
  • SKIP 2
  • 8 lessons

The 'design from scratch' narrative + core patterns + fixtures + CI/parallel. STOP before test platforms, custom runners. Over-patterning is a negative signal.

  • MUST 19
  • SHOULD 10
  • STRETCH 3
  • SKIP 2
  • 14 lessons
Topic 5Python + Pytestnot started
42 items

Full OOP/decorators/generators + deep Pytest; DSA ONLY Easy→comfortable-Medium (0/90 JDs ask Hard). STOP before LeetCode Hard/DP, metaclasses, CPython internals.

  • MUST 25
  • SHOULD 14
  • STRETCH 2
  • SKIP 1
  • 4 lessons

Application-level TS to read/write/maintain a Playwright suite (async/await is the real gap). Unlocks the 13% TS-only roles. STOP before compiler-level type gymnastics.

  • MUST 13
  • SHOULD 14
  • STRETCH 7
  • SKIP 3
  • 7 lessons
Topic 7CI/CDnot started
41 items

Own the TEST STAGE — 'I consume the platform, I own the test stage.' Jenkins slightly outranks GH Actions in our pool. STOP before Jenkins admin, K8s ops, Terraform, deployment mechanics.

  • MUST 14
  • SHOULD 15
  • STRETCH 9
  • SKIP 3
  • 17 lessons
Topic 8Performance / Loadnot started
32 items

Competent user of ONE tool (k6) with a CI gate + results literacy. SHOULD-level differentiator, not a MUST gate. STOP before LoadRunner, JVM tuning, capacity modelling.

  • MUST 11
  • SHOULD 11
  • STRETCH 4
  • SKIP 6
  • 9 lessons
Topic 9SQL + Data/ETLnot started
43 items

Up to window functions + CTEs; SQL-as-test-oracle. A hard MUST (~35/50 JDs), not a quiet filter; ETL reconciliation is a MUST for Tekion/Cognizant. STOP before query tuning, DBA ops, pipeline authoring.

  • MUST 16
  • SHOULD 11
  • STRETCH 12
  • SKIP 4
  • 6 lessons
Topic 10Cloud / Infranot started
29 items

Fluent USER to run/debug tests, never owner. Docker highest ROI; K8s = actually-run kubectl (Apply). STOP before certs, Terraform/IaC, cluster ops.

  • MUST 9
  • SHOULD 12
  • STRETCH 3
  • SKIP 5
  • 11 lessons
Topic 11AI/LLM Testingnot started
45 items

Build & own an LLM eval harness in CI (Part A) + AI-assisted testing is now table-stakes (Part B). Capability-first vocabulary. STOP before training/fine-tuning, transformer internals, MLOps.

  • MUST 16
  • SHOULD 20
  • STRETCH 6
  • SKIP 3
  • 15 lessons
Write to me

If any of this is something you can help with, or you think I have got it wrong, write and tell me. I read them all myself.

Dhanunjaya M.Dhanunjaya M
tvsdhanan009@gmail.com →

One address, no form, no list to join. I am not asking for money, and there is nothing set up here that could take any.