khanaitechnologies

VERIX AI is a Windows-native AI testing agent designed to make desktop application testing more accessible, intelligent, and verifiable. Traditional automation often depends on manually scripted interactions and application-specific test frameworks. VERIX AI addresses this by combining Meta Muse-powered reasoning with local Windows UI Automation, human-reviewed execution plans, and independent verification. The system can inspect desktop interfaces, identify accessible controls, execute explicitly authorized actions, and collect screenshots and structured evidence. Its workflow follows Observe → Plan → Review → Act → Verify, with safety controls, execution limits, and replayable records. Our demonstration showcases the VERIX AI interface and local application testing workflows using Windows Calculator and FreeCAD. Technology: Python, Windows UI Automation (UIA), Meta Model API / Muse Spark, local multi-agent orchestration, and evidence-based verification. Goal: Advance reliable, transparent AI-driven desktop testing while keeping users in control of application interactions. Current limitation: Meta Muse reasoning and local execution have been tested separately; a complete Muse-generated executable plan through verified native execution has not yet been independently established.