AI-powered QA
Testing smarter with AI, with a human in the loop.
I build AI tools that turn requirements and code into test plans, test cases and Playwright suites. Then I review every result, because AI output is a fast first draft, not a verdict.
- specwright run --prd login_prd.md --scope regression
- Stage 1 · Planner — IEEE-829 test plan ready
- Stage 2 · Case writer — 24 cases, sized per requirement
- Scope gate · Regression — P0 + P1 selected
- Stage 3 · Automator — Playwright suite with Page Objects
- Coverage check — REQ-4 needs a human look
- Reviewed by Rafi — ready for CI
- AI test planning agentsAn instruction file in VS Code that reads the requirement and the codebase together, then returns an impact analysis, test plan, test cases and risk register.
- LLM-generated test suitesSpecwright turns a PRD into an IEEE-829 test plan, AI-sized test cases and a Playwright TypeScript suite, all on free LLMs.
- Think first, AI secondI map the risky scenarios myself (session timeouts, token expiry, concurrent logins), then use AI to widen coverage and review every case it writes.
- Smarter regressionAI-assisted failure pattern analysis, coverage gap detection and regression prioritisation, so the right tests run first.
- Groq · Llama 3.3
- Gemini
- Claude
- Ollama
- Playwright
- Python
- Streamlit
Our Blog
Stay On Top Of The
Latest News
The Markdown File That Found a Security Hole Four Engineers Missed
How to build an AI test planning agent in VS Code that reads your requirement and your codebase together —…
Don’t Blindly Trust AI-Generated Test Cases: A QA Lesson I Learned the Hard Way
A few weeks ago, I was sitting in front of my laptop, staring at a critical login flow I was…
Writing Code That Actually Matters: A Developer’s Quiet Role in the SDGs
Amina didn’t think about global goals or digital transformation when her online class stopped loading. She just stared at her…