I care about building AI workflow systems that are understandable, maintainable, and reliable under real use.
Recently, I have been contributing to open-source AI infrastructure, with a focus on backend behavior, scheduler reliability, MCP/tooling boundaries, and review-ready fixes.
current focus:
- AI agents and workflow orchestration
- backend architecture and reliability
- open-source contribution workflows
- debugging, tests, and maintainable patches

