Solutions Architect, building enterprise-scale systems with Global FinTech. Now writing about the parts nobody puts in the architecture diagram.
I have spent over a decade architecting systems, which has taught me that the interesting problems are rarely the technical ones - they are the trade-offs, the governance questions, and the "why did we actually build it this way" conversations. I write about enterprise architecture, AI adoption in regulated industries, and the occasional opinion on developer tooling.
Markdown supported · ⌘↵ to post
Ahmet Özel
Model selection should be driven by task risk and acceptance criteria, not perceived intelligence alone. A useful routing policy records the task class, latency budget, expected output schema, and fallback. Then the team can compare accepted quality and total retry cost rather than choosing models by anecdote.