Blog

Engineering write-ups, product updates and a recurring AI Ops report — the three content pillars of §8.3.4.

EngineeringHow a flow is actually built — the blocks, the ports, and the trade-offs between them.
ProductWhat shipped, and why it is shaped the way it is.
AI OpsWhat breaks in a production AI application, and which of the Quality modules catches it.

Docs, the blog, and a first look at the Marketplace: Three new public surfaces: a docs site with a runnable quickstart, this blog, and a teaser of the ten categories the template Marketplace opens with.
State of AI Ops: what breaks first when you ship an LLM feature: Three failure modes show up before any others, in roughly this order: a silent prompt regression, retrieval drift nobody measured, and a bill that points at a month instead of a feature.
Building a RAG flow with Vectorize and Workers AI: Four blocks, one graph: embed the question, query the index, assemble the prompt, generate the answer. What each step actually validates before the next one runs.