NnRouter.aiinnrouter.hashnode.dev·3h ago · 3 min readReserve Before Dispatch: Budgets That Fire Before InferenceMost LLM budget tooling is a notification system. The request runs, the provider bills, the dashboard updates, and at some threshold an alert informs you that money has been spent. By the time a human00
NnRouter.aiinnrouter.hashnode.dev·6h ago · 3 min readRouting Is a Configuration Change: Putting the Model Choice Where It BelongsSomewhere in your codebase there is a model name, hardcoded or configured once, chosen when the feature was built. At the time it was the right choice — probably the best model available. The problem 00
NnRouter.aiinnrouter.hashnode.dev·7h ago · 2 min readFive Steps Every Request Takes: Reading nRouter's Introduction as an Architecture DiagramDocumentation introductions are usually marketing in disguise. nRouter's Introduction page is better read as an architecture diagram written in prose — because its central content is a sequence: the f00
NnRouter.aiinnrouter.hashnode.dev·23h ago · 3 min readThe ai-proxy Translation Table: Five Kong Plugins, Five Different HomesThe most useful page in any gateway evaluation is the translation table: not what each product has, but where each job you currently run goes. For teams on Kong, the jobs live in the plugin chain hang00
NnRouter.aiinnrouter.hashnode.dev·1d ago · 4 min readPre-Call Enforcement: Why Governance Has to Run Before the Provider CallPre-Call Enforcement: Why Governance Has to Run Before the Provider Call Every governance control has a position. It can sit after the request — in a dashboard, a weekly report, a post-incident review00
