I'm building SYSTEM-ENGINEERING to explore a question I believe is becoming increasingly important: how do we make AI-assisted software development more structured, verifiable, and maintainable?

The project explores agent orchestration, explicit policies, validation workflows, and persistent project context.

I'm sharing it with the community at an early stage because I'd value honest technical feedback more than vanity metrics.

If you work with AI coding assistants, agents, or developer infrastructure, I'd love to hear your perspective:

What is the biggest engineering challenge you face when working with AI agents?
Which parts of this approach would you challenge or redesign?
What would you need to see before using a system like this in a real workflow?

Explore the project: https://github.com/SHEMASOFT/SYSTEM-ENGINEERING

Feedback, critical questions, bug reports, and contributions are welcome. Thanks for taking a look!