EvaQA: An Autonomous AI-Powered QA Testing Platform with Natural Language Driven Visual Test Orchestration

Software quality assurance remains one of the most resource-intensive phases of the software development lifecycle. This paper presents EvaQA, an autonomous desktop-based QA testing platform that enables non-technical users to author, execute, and report on automated tests through plain-English natural language descriptions alone. EvaQA generates executable Playwright test scripts via a locally hosted LLM (Ollama), orchestrates execution with real-time animated flowchart visualisation using React Flow, and resolves AI selector hallucination through a proprietary DOM Context Injection Pipeline. The platform supports multi-role collaborative sessions, automated PDF report generation, and screenshot evidence capture — all in a fully offline-first, privacy-preserving architecture. Evaluated across 47 web applications (235 scenarios), EvaQA achieved 91.3% test script generation accuracy with an average test plan generation time of 8.2 seconds.