Ömer ÇelikLead Developer | Enterprise systems architecture · Governed AI in software delivery
I build: enterprise .NET systems you can prove
I design and build enterprise systems on .NET: order management, workflow orchestration and integration. And I work on how a team can use AI in delivery while people still decide what reaches production.
Practice
Six areas I work in, each with something you can open.
Work
Case studies — context, decision, outcome. Client names withheld.
Open source
The code behind the practice, on GitHub.
AI-native, spec-driven enterprise .NET accelerator — composable libraries, templates, AI skills and guardrails on a paved golden path. Not a framework.
Deterministic spec lint for manifest-driven golden paths — validates cross-artifact invariants and detects manifest-vs-repo drift, served over MCP. LLMs call it; it never calls an LLM.
Writing
Essays with diagrams and demos. One idea each.
Labs
Ideas you can try in the browser. Mock data, nothing leaves the page. The same demos open inside the essays.
Spec drift playground
Change one field in a manifest and watch a deterministic gate catch it.
specOpenTest pyramid, after AI
Drag the cost of writing code toward zero and see where trust has to move.
qualityOpenWorkflow stepper
Step one product order through its process, watch the events it emits, and try to cancel it midway.
orchestrationOpen