Agentic QA Framework
Overview
AI-powered QA framework that autonomously generates Playwright end-to-end tests and runs them on a Telegram chat command.
Responsibilities
- Designed and built the agentic framework around an OpenClaw agent that inspects the app and authors Playwright specs.
- Implemented Telegram-triggered test runs so anyone on the team can start a suite without CI access.
- Committed generated specs as reviewable source and reported results, traces and screenshots back to the Telegram thread.
Result
- 60% reduction in end-to-end testing time
- Test authoring no longer competes with feature delivery for engineer time
- Any team member can trigger a full regression run without CI access