Beneath the Claim

Before the models, the substrate: why source-control hygiene and CI/CD maturity decide whether enterprise AI ships or stalls, and a scorecard for measuring it.

AI readiness is not a tool you buy or a checklist you complete. It is the compound outcome of three decisions: how you structure code, how change flows, and how you deliver, and why the weakest of the three caps the...

What breaks, and what you have to rebuild, when a respected technology hub grows from 32 to 220 in three years: onboarding, org structure, the operating model, and the founders you cannot afford to burn out.

Every board is being asked to approve AI spend. The harder question no one puts on the slide: AI multiplies the engineering organisation you already have, so are you funding a multiplier on strength, or on fragmentation...

Long before AI multiplies anything, a fragmented engineering estate levies a quiet annual tax: in context-switching, duplicated pipelines, slow onboarding and a security bill that scales with sprawl. The consolidation...

For thirty years we governed the humans who touch the codebase and trusted the tools they held. An agent that opens a pull request and pushes to a branch is not a tool. It is an actor, and it needs an identity, an...

You can't decree AI readiness into an organisation with strong team autonomy. You pilot with the willing, publish the delta, and let demonstrated value manufacture the champions who pull the rest along.

Ask the wiki and your organisation has a branching standard. Ask the repositories and the truth comes back: a dominant flow nobody decided, a long tail with no model at all, and branch lifetimes nobody governs, just as...

The monorepo-versus-polyrepo war is back, and both camps are arguing about the wrong variable. What an AI agent, and a new engineer, can actually use is contextual coherence, and for a brownfield enterprise the road...

Almost no AI forecast can be graded, because almost none carries a date, a number and a company staked on it. One did. Scoring it two years on shows which layer of a vendor pitch is worth your money, and which one is...

The same model scored 0.637 and 0.488 on the same benchmark because two teams spelled the answer differently. Every AI number you have been shown is a scoring decision somebody made, and the one rule that encodes what...

A company that sells AI adoption software surveyed people already using AI, folded seven different behaviours into one word, and published the cure. The word travelled. The reaction it is producing costs more than the...

Asking for the source is the instinct that feels like diligence. In the two best documented cases available it fails: the source published several different values, and the caveat was dropped by one of the people who...

In a company that already exists, the first move toward AI readiness isn't a migration. It's a measurement baseline. You cannot strengthen what you haven't instrumented, and you cannot prove the change paid off if you...

Your software stack is not infrastructure you own. It is a dependency you rent from firms that answer to a government, and that government just proved it will pull the switch.