Architecture
5 articles tagged architecture.
Model Routing: Stop Paying Frontier Prices for Formatting
— Picking one model for your whole product is a pricing decision disguised as an architecture. Routing per step is cheaper, faster, and easier to govern.
The Semantic Layer Is What Makes an LLM Useful on Your Data
— An LLM with database access is not an analyst. The semantic layer is where entities, metrics, and permissions get defined once and honored every time.
Disposable Software: When the Interface Exists for Ninety Seconds
— AI can now generate a single-use view for one task and discard it. That is genuinely useful, and it quietly breaks three assumptions your app depends on.
MCP and the Governed Tool Hub: Giving Agents Safe Superpowers
— The Model Context Protocol lets agents use real tools. A registry, a gate pipeline, and workspace-minted tokens are what keep that power from becoming a liability.
The Agentic Mesh: Coordinating Multiple AI Agents on One Product
— One monolithic agent buckles under a real product's variety. A mesh of specialists that delegate and hand off scales better, and stays legible.
← All Field Notes