A GAP prototype uses Claude Code to generate test cases directly from user stories and execute them in a real browser via Playwright. The system automatically drafts defect tickets for failed runs, accelerating the quality assurance process while maintaining human oversight.
Building an AI-Driven QA Pipeline with Claude Code Read More »
Two GAP engineers demonstrate an agentic pipeline built on Claude Code that plans, executes and closes Azure DevOps tickets within a client monorepo. Specialized AI agents generate implementation plans, write code and run validation tests, making the work traceable from plan to commit.
Ticket to Commit: An Agentic Pipeline Built on Claude Code Read More »
A GAP engineer demonstrates Nextly, an AI portal assistant that provides guided in-app navigation without reading page HTML. The standalone widget uses PDF guides and element-ID mapping to answer user queries, automatically scrolling and highlighting the exact interface steps required.
Guided In-App Navigation With an AI Portal Assistant Read More »
A GAP QA engineer demonstrates SWAT, a prototype agent-based testing framework. An orchestrator coordinates specialized agents to generate and execute Playwright test cases and maintain visual baselines, condensing a month of hand-coded automation into under two hours.
Agentic QA: Autonomous Web Test Generation With SWAT Read More »
Decentralized AI tools often lead to fragmented engineering workflows. This prototype demonstration introduces a centralized, governed agent platform designed to standardize these processes. The walkthrough illustrates how specialized AI agents can generate Azure DevOps tickets from Figma designs and draft QA test cases directly from pull requests, while keeping human engineers in control of final
Building a Governed Agent Platform for Engineering Teams Read More »
