Medium

Amrutalohabare

Pinned

In

The Ai Studio

by

Amrutalohabare

·

Jun 1

Using AI to Find Coverage Gaps in Your Playwright Test Suite

You can’t see your blind spots from inside your own test suite. Here’s how I use AI to find them.

Using AI to Find Coverage Gaps in Your Playwright Test Suite

Pinned

Amrutalohabare

·

Apr 5

Migrating from Selenium to Playwright: Real Challenges Teams Face (and What Actually Works)

Migration from Selenium to Playwright is often presented as a straightforward upgrade.

Migrating from Selenium to Playwright: Real Challenges Teams Face (and What Actually Works)

Amrutalohabare

·

6h ago

I Let an AI Judge My AI. It Disagreed With Itself.

Learning to test AI, Session 8 (LLM-as-judge & custom rubrics): two near-identical answers, opposite verdicts, and why the rubric is part…

I Let an AI Judge My AI. It Disagreed With Itself.

Amrutalohabare

·

6d ago

Two Tests Failed. Only One Was the AI’s Fault.

Learning to test AI, Session 7 (your first automated eval with promptfoo): 15 tests, two models, one command, and why the score was the…

Two Tests Failed. Only One Was the AI’s Fault.

Amrutalohabare

·

Nov 1

My AI Returned Perfect JSON. Then It Added an Essay.

Learning to test AI, Session 6 (structured outputs & output validation): the same prompt, two runs, two formats, and why valid JSON isn’t…

My AI Returned Perfect JSON. Then It Added an Essay.

Amrutalohabare

·

Oct 26

My Chatbot’s Most Dangerous Bug Was Its Most Helpful Answer

Learning to test AI, Session 5 (prompt failure classification): I tested my HR bot against all 8 categories of prompt failure. The critical…

My Chatbot’s Most Dangerous Bug Was Its Most Helpful Answer

Amrutalohabare

·

Oct 25

The Best Thing My AI Skills Do Is Check Before They Act

Three skills, and the step that saves me the most time isn’t the writing.

Amrutalohabare

·

Oct 17

I Told My HR Bot a Lie. It Emailed the Whole Company.

Learning to test AI, Session 4: I wrote a “complete” system prompt for an HR bot, then broke it with an unverified claim and a little…

I Told My HR Bot a Lie. It Emailed the Whole Company.

Amrutalohabare

·

Oct 14

AI Tests a Moment. Your Users Live in Time.

Every generated test starts clean and finishes in two seconds. Real users don’t work like that.

AI Tests a Moment. Your Users Live in Time.

Amrutalohabare

·

Oct 13

Every AI Chatbot Looks the Same. The Bugs Don’t.

Learning to test AI, Session 3: Simple LLM, RAG or agent, and why the type decides what you test first.

Every AI Chatbot Looks the Same. The Bugs Don’t.

Amrutalohabare

Amrutalohabare

Remote Work | Freelancing | ISTQB CTFL | Playwright | Selenium | MCP | Agentic-AI | Agile | Automation Test Engineer | QAOps

Following

Help

Status

About

Careers

Press

Blog

Store

Privacy

Rules

Terms

Text to speech