Senior Quality Engineer · Test Automation · AI / ML QA
Issue 01 · Volume 01
A working definition
I build the test automation that ships trustworthy software — across web, mobile, API, and AI. The work sits at the intersection of test engineering, product quality, and modern delivery — end-to-end QA strategies built from scratch, frameworks that survive contact with real teams, and a culture where quality is a product feature.
A working reel · 2020 — 2026
About · 01
SDET at noah.com working on blockchain-integrated products. Previously QA across fintech, marketplace, insurance, and edtech — through direct engagements and freelance platforms.
I'm Pankaj — a QA engineer and SDET based in Dubai. For the last 15+ years I've been the person teams call when they want their software to actually work in production, not just on staging.
My work sits at the intersection of test engineering, product quality, and modern delivery. I design end-to-end QA strategies from scratch, build the automation frameworks that hold them up, and ship alongside the engineers — Playwright and TypeScript when I can choose, Cypress, Selenium, WebdriverIO, or Appium when the stack calls for it.
Lately I've been deep in AI/ML QA: testing LLM applications, agent workflows, and the brittle seams where AI features meet real users. I'm equally comfortable with a GraphQL query, a JMeter load profile, or a flaky mobile test on BrowserStack.
I care about clean execution: small, fast feedback loops, frameworks that survive contact with real teams, and a culture where quality is a product feature, not an afterthought.
Expertise · 02
I work where testing meets product. The work splits naturally into four overlapping practices — the seams between them are where the value lives.
Read more
End-to-end frameworks in Playwright, Cypress, Selenium, and WebdriverIO. Page-object discipline, fixtures that survive, CI wired for parallel runs.
Playwright · Cypress · WDIO · Selenium
Read more
Strategy and tooling for LLM applications, agent workflows, and prompt quality. Where AI features meet real users — the brittle seams.
Dify · Ollama · Cursor · Eval frameworks
Read more
Contract testing, schema validation, load profiles, and the kind of slow-leak bugs that show up the day after launch.
Pytest · SuperTest · Locust · JMeter
Read more
Cross-browser, cross-device, real-device on BrowserStack and Bitrise. The boring work that decides whether a release ships.
Appium · React Native · Flutter · Cypress
Career · 03
The arc below is the same one most senior QA leads walk — but with the AI/ML inflection added in the last few years. Each role taught a different part of the craft.
A career · 2010 — Present
Originally a manual tester, I began writing automation at a small agency where the only QA tool that worked was the one I built myself. That early experience shaped how I work today: the framework is the product, not an afterthought. The career since has been a long, deliberate compounding — from Selenium + Java in 2013, to Cypress + JS in 2019, to Playwright + TypeScript in 2023, to the AI/ML work that dominates my current focus.
2025 — Present
Full-time
noah.com · Remote
Owning end-to-end test automation for a fintech platform with blockchain-integrated features and AI-driven workflows.
2024 — 2024
Contract
Contract Role · Remote
Embedded QA for an AI/ML product team building LLM applications, agents, and orchestration workflows.
2022 — 2024
Contract
Upwork core team · Remote
Domain QA specialist across multiple product launches, leading both manual and automation workstreams.
2022 — 2024
Contract
Grey Orange Inc. · Remote
Central QA contact for multiple clients, building test automation for robotics-platform-adjacent products.
2022 — 2023
Contract
Chef Booking Platform · Remote
End-to-end QA for a marketplace product — manual testing, no-code UI automation, and PR-level defect prevention.
2022 — 2022
Contract
Prudential · Remote
QA on a 360-degree feedback tool revamp for a 14,000-employee Asian insurance provider.
2020 — 2022
Contract
GrowthDay · Remote
Owned QA across web, mobile (Android & iOS), and back-end — built a 500+ case API automation suite.
Toptal · 04
One of the world's most selective talent networks. Book a senior QA engineer the same way you'd book a top-tier consultant — without the recruiting overhead.
Why Toptal · A short note
I've joined the Toptal network for the same reason clients use Toptal to find me: the signal-to-noise ratio is the best in the industry. The screening is what makes the marketplace work, and the trial period is what makes it risk-free.
Toptal screens every engineer through a multi-stage process. The Top 3% figure is theirs, not mine.
No sourcing, no recruiting, no HR paperwork. You get a matched expert and a signed SOW inside a week.
If the first two weeks don't click, Toptal refunds the engagement. I've never had to use it.
NDAs, IP assignment, and invoicing are handled by Toptal. We both focus on the work.
Selected Engagements · 05
A working list of the teams I've shipped with. Most engagements started as a contract and ended as a long-term relationship.
Table 02 · Selected engagements · 2020 — 2026
Selected Work · 06
Every project below is on GitHub — built around a problem I actually wanted to solve. Sorted with the current focus on top.
No.
01
Automation Framework · 2024
Production-grade end-to-end test framework for Sauce Demo using Playwright + TypeScript with the Page Object Model.
No.
02
AI Tool · 2025
Generate realistic, scenario-aware test data locally using Ollama — no PII leakage, no cloud round-trips.
No.
03
Automation Framework · 2024
Playwright + TypeScript framework purpose-built for automating Excel Online's Today() function and formula behaviour.
No.
04
Automation Framework · 2023
Python + Selenium framework for the BookCart reference app — clean structure, pytest reporting, and POM patterns.
No.
05
Test Utility · 2024
Side-by-side comparison of how Selenium, Cypress, and Playwright handle Shadow DOM piercing — with runnable code.
No.
06
Automation Framework · 2024
Playwright automation suite for a Luma-style event flow.
No.
07
Automation Framework · 2023
A clean Cypress JavaScript starter for greenfield test automation projects.
No.
08
Automation Framework · 2023
Minimal, no-frills Python + Selenium test automation starter.
No.
09
Mobile · 2023
Appium tests for Android with WebdriverIO as the client — ready to point at your APK.
No.
10
Scraper · 2022
Lightweight Python scraper that turns Yellow Pages search results into a clean Excel file.
Stack · 07
Tools I reach for under deadline pressure — and the ones I keep around because they pay off on the long arc.
Frameworks I reach for, in order of how often I open them.
Comfortable shipping production code — not just tests.
Where the products I test actually live.
Validating contracts and load behaviour before users find the seams.
Pipelines and infra I integrate with, not just touch.
Where my current focus sits — testing LLM apps and AI workflows.
The glue that holds a test stack together.
Approach · 08
The rules I work by. They're not a manifesto — they're a pattern that survived contact with dozens of teams and products.
I build the test system, not the test cases. A good framework grows with the team; a good test case is just today's expectation captured. Invest in the system.
The cost of a defect in production is not a line item — it's a trust withdrawal. I treat testability as a first-class design constraint, not a downstream concern.
I prefer the tools I can debug at 2am. Playwright, TypeScript, Python, GitHub Actions, Postgres. The novelty budget goes to the AI work, not the rest.
CI on every PR, nightly regression, weekly exploratory. The compounding cost of slow feedback is the silent killer of quality programs.
I work inside the team. Same Slack, same standups, same off-site. The QA lead who shows up is the one whose fixes get merged.
The interesting bugs are at the contract boundaries — third-party APIs, AI outputs, async queues. That's where the most leverage is, and where I focus.
Lens · 09
A snapshot of the things I'm reading, building, and working on this season. Updated roughly monthly.
Colophon · 10
A small editorial aside — the human parts of the work that don't show up in a job description.
Origin
Assam, India · 1988
Home
Dubai, UAE · 2018 —
Languages
English · Hindi · Assamese
Coffee
Strong, no sugar
Family
Married · two kids · Eva & Neev
Side project
AI test-data generator · Ollama · Open source
Dispatches · 11
I write occasionally — about test engineering, AI / LLM QA, product thinking, and the long arc of a QA career. No roundups.
01
May 18, 2026
Software Engineering
A field guide to the mental models I use to ship software that doesn't quietly break in production — from test pyramid shape to on-call trust.
02
Apr 22, 2026
Product Thinking
The thing that separates a senior engineer from a senior code-monkey isn't raw skill — it's caring about what the code is for.
03
Mar 10, 2026
AI Tools
Cursor, Copilot, and the new shape of the SDET workflow — what's real, what's hype, and how to integrate AI tools without breaking the things that make a test suite trustworthy.
Contact · 12
Open to senior SDET, QA lead, and test-automation consulting engagements. Drop a line — happy to chat about scope, timelines, and the actual shape of the work.