Latest insights

AI-assisted coding is becoming table stakes. AI systems engineering is becoming the real differentiator. Here is the difference, and why it matters.

Tool-using agents need an explicit map of what they can read, write, mutate, escalate, and never touch.

Before you tune retrieval, prove the question can be answered from the source corpus, with a citation a human would accept.

Production AI fails in ways ordinary app runbooks do not cover. The operating plan has to include quality drift, retrieval failure, model outages, cost spikes, and human escalation.

Prompt caching is no longer a performance optimization. It is an architectural constraint that decides whether a long-running agent is economic to operate.

The Model Context Protocol is moving from a developer convenience to the production interface between agents and your systems. Here is what changes when you treat it that way.

A small verifier model sitting between the frontier model and the side-effect boundary is the most useful piece of agent architecture nobody is shipping yet.

How a working eval harness picks the model — and how often the answer is not the frontier one the team came in expecting.

A PoC with no path to production hides the hard decisions. A week-2 user forces them into the open while the architecture is still cheap to change.

Why we put a real user inside the system in week 2, and the kinds of architecture decisions that get rewritten when we do.

How the discovery eval is supposed to work — including the two times we walked away and refunded the fee.

An evaluation harness is a product. Ship it like one — versioned, owned, and instrumented.