
How three Java Champions put GitHub Copilot to work
How are experienced Java developers using AI coding tools beyond autocomplete? In my latest GitHub Blog article, I interviewed Java Champions Vlad Mihalcea, Patrick Baumgartner, and Frank Delporte about their workflows.
Vlad describes accelerating feature development while preserving existing coding practices. Patrick shares how he approaches agentic workflows with clear requirements and task-specific verification. Frank walks through a Pi4J investigation where Copilot traced a hardware-input bug to virtual-thread pinning in native calls.
Across these different experiences, the same practical advice emerges: give the agent useful context, ask for evidence, review its output, and remain responsible for what ships.
Found a mistake, or something to add? Edit this page on GitHub








