agentgateway content-based routing explained and demo'd

🚀Content-based routing in agentgateway — the smartest way to route LLM requests based on what's actually inside the prompt or request body. In this video I break down exactly how content-based routing works in agentgateway, why it matters for cost control, performance, and multi-model strategies, and then I demo it live end-to-end. ### What you'll learn: - What content-based routing actually is (and why header/path routing isn't enough) - How AgentGateway extracts fields from the JSON request body (like `model` name, prompt length, or custom metadata) - Step-by-step configuration using route-level transformations + header matching - Real-world patterns: cost-optimized routing, model-family separation, fallback strategies, and more - Live demo: sending requests and watching traffic intelligently route in real time Perfect for anyone building production LLM infrastructure, multi-agent systems, or cost-efficient AI platforms on Kubernetes (or standalone). ⏱️ Timestamps: 00:00 - Why content-based routing changes everything 01:45 - How AgentGateway implements it (transformations + headers) 04:20 - Full config walkthrough (YAML explained) 08:15 - Live demo: routing by model name & prompt content 14:30 - Advanced patterns & best practices 18:45 - Common gotchas & troubleshooting 21:10 - Summary + next steps 🔗 Resources & Links: • AgentGateway docs (content-based routing): https://agentgateway.dev/docs/ • GitHub / Repo: https://github.com/sebastianmaniak/agentgateway-demos/tree/main/05-content-based 💬 Drop your questions in the comments — happy to help debug your routing config! 👍 Like if this helped you level up your LLM infrastructure 🔔 Subscribe for more agentgateway tutorials, real demos, and production AI patterns every week #agentgateway #LLMRouting #ContentBasedRouting #KubernetesGateway #LLMInfrastructure #AIAgents #MultiLLM #OpenAI #Anthropic #xAI #AIObservability #DevOps #Kubernetes