Nima Mahdavi

Visual Change Detector is a visual regression testing tool that automatically detects UI changes across builds in web applications. I designed and developed this project to address the limitations of traditional functional tests, which often miss visual inconsistencies such as layout shifts, styling issues, and unintended UI regressions.

The tool uses browser automation to capture screenshots of web pages in a consistent environment and compares them using image techniques to identify pixel-level differences. Detected changes are highlighted and reported in a clear, actionable format, enabling faster investigation and reducing manual visual validation efforts.

I handled the full design and implementation of the solution, including test execution flow, screenshot management, comparison logic, and result reporting. This project demonstrates my expertise in QA automation, visual regression testing, and practical tooling that improves test coverage, reliability, and release confidence in modern web applications.

I joined Govin B.V., an early-stage Legaltech and AI compliance startup (ISO 27001-certified, GDPR-compliant platform), at a critical growth phase when no formal QA structure existed. I built the entire quality assurance process from the ground up — starting with manual test case design, defining testing workflows, and establishing defect management practices using Jira and Azure DevOps. I trained and mentored a team of three engineers in test case design, execution, and reporting standards. Given the compliance-sensitive nature of the platform, manual testing remained central throughout — smoke tests, new feature validations, and critical regression scenarios were always verified manually before automation sign-off. Once the manual foundation was solid, progressively introduced test automation using Playwright, Selenium, and Robot Framework, and extended mobile coverage with Appium. This phased approach allowed the startup to achieve reliable release quality within budget constraints, significantly improving regression coverage and product stability across releases.