Visual QA Checker for BilaUnwan.pk
An AI-powered visual quality assurance tool that detects layout shifts, broken links, and UI regressions across publishing workflows—saving time and ensuring a polished user experience.
Visual QA Checker for BilaUnwan.pk
A production-grade, automated QA tool built to monitor and detect visual anomalies, broken links, and layout drifts across the BilaUnwan.pk publishing platform. This solution was tailored for a fast-paced content environment with minimal manual QA.
Business Challenge
- Frequent layout shifts across devices and browsers
- Manual testing was slow and error-prone
- Content updates caused hidden frontend regressions
- Dead links and UX bugs affected reader trust
Solution Highlights
- Automated screenshot crawling of live pages
- ML-based detection of visual inconsistencies
- Real-time alerts via email, Slack, Telegram
- Dead link reporting dashboard
- Retrainable pipeline with issue labeling
Impact
- Reduced manual QA time by 80%
- Rapid bug discovery after each update
- Cross-device and cross-browser coverage
- Enabled confidence in scaling content campaigns
Tech Stack
Frameworks: FastAPI, React
DevOps: Docker, GitHub Actions
ML Tools: MLFlow, Custom CV Models
Others: Python, Puppeteer, Slack API
Why It Matters
FursanStudio's solution helped a grassroots publishing platform implement enterprise-grade QA — affordably and flexibly — without vendor lock-in or technical overhead.
Project information
- Category: AI / QA Automation
- Client: BilaUnwan.pk
- Project date: April 2025
- Report: Download PDF
- Request Demo