You don't have an AI problem. You have a transformation problem.

The question is no longer whether AI works. It’s whether your team can transform in time to use it.

Something shifted in November. I've spent the months since in back-to-back conversations with customers, partners, and engineering leaders, and the same thing keeps surfacing. The urgency is different. Not in a hype-cycle way. More like a quiet recognition the window is narrower than most orgs thought, and that the decisions getting made right now about structure, tooling, and investment are going to be hard to undo.

What I'm not hearing anymore is "does this work." What I am hearing is "how fast do we need to move" and "what happens if we don't."

The organizations falling behind aren't losing because the AI doesn't work. They built their org structure, their decision-making, and their spending philosophy for a different era: approval processes, per-seat budget debates, “one more POC” to decide, “what does the analyst ranking say.”

The startups being founded right now don't have any of that to unwind. They're AI-native by default, smaller, and their cost structure looks nothing like an established engineering org's. Every quarter a larger company spends in deliberation is a quarter that gap gets harder to close. Organizational habits compound just like the technology does.

Rajeev Rajan, CTO of Atlassian, bought a personal laptop over the holidays because corporate IT blocked him from installing Claude Code. When Thomas Dohmke, former CEO of GitHub, heard this, his response to startup founders was immediate: "When an investor asks how you're preventing the incumbent from doing the same thing you're doing, just tell them the CTO of Atlassian had to buy a laptop on his own money to start coding."

It's a funny story. It's also a precise description of the problem. Resources aren't the constraint. Structure is.

When I talk to engineering leaders right now, I find them in roughly three places. Red: still in deliberation, still proving the case. Yellow: adopted, but hitting new bottlenecks. Green: committed, agents running, thinking about throughput. The distribution is not what most people expect.

Most engineering orgs are in red. The ones pulling ahead have stopped debating it.

When agents are running at scale, alignment becomes the binding constraint. Agents do what they're told. If the org doesn't have genuine shared clarity on what it's building and why, you end up with hordes of capable agents executing confidently in different directions. That's harder to see than the old kind of dysfunction and much harder to fix, because the output volume is high and speed can feel like progress until it reveals itself as sprawl.

The orgs that will most successfully navigate green territory are the ones that do work that doesn't look like AI work at all. They will be explicit about strategy and principles before they need to. They will create a shared context that lets all functions and agents operate with less friction and more confidence. Not a culture deck. Something more like a spec: specific enough to actually guide decisions.

This isn't new thinking. The best-run companies have always operated this way. What's changed is the cost of skipping it. Attrition accelerates when engineers feel like their org is behind. Costs compound when agents are running without direction. And the gap with companies that started earlier widens in ways that are genuinely difficult to reverse.

I've had some version of this conversation dozens of times in the past few months. The organizations that started moving early, even imperfectly, are pulling ahead. The ones waiting for the right moment are finding that the right moment has a cost.

Where are you today? Red, yellow, or green?

We are learning what this journey looks like because we're going through the transformation ourselves. Our own engineering org is going through the same phases, hitting the same friction points, and figuring out the same organizational questions our customers are. When we talk about what it takes to move from red to yellow to green, or what breaks down when agents scale without alignment, we're not speaking theoretically. That's what gives us something real to offer beyond the tooling. We're mapping this terrain, we're still learning, and we can help you through your own journey.

If you're working through this and want to compare notes, reach out.