Insights.
The papers we have written and what they taught us. Everything here comes from work that shipped.
Before you give AI the keys
Telling an agent not to do something is not the same as preventing it. This paper looks at the governance layer that needs to be in place before rollout: what an agent is allowed to do, what it is allowed to spend, and what it is allowed to reach.
Knowing your company brain actually works
A company brain can look healthy and still be quietly wrong. This paper closes the series with three checks: whether the brain finds the right pages, whether the agents maintaining it kept the meaning, and whether every page still earns its place.
The maintenance layer most teams forget to build
Most company brains drift, go stale and lose the trust they were built to earn. This paper is about the maintenance layer that keeps them worth using.
Your company brain is probably an archive
Building a company brain is the easy part. Keeping it clean, tight, and actually useful is the discipline most teams skip, and it decides whether anyone keeps using it.
How we build fast AI assistants
Five principles for speed, cost, and reliability in production AI systems.
Change how
you operate.
Tell us where you’re stuck. If operational AI can change the outcome we’ll show you how, and if it can’t, we’ll say so.
Talk to us