How we engage.

What we have instead is a sequence. Three points where you commit a little, learn a lot, and decide whether the next step is worth taking. Each one stands on its own. Each one is built to be easy to walk away from. You are never buying a pipeline. You are buying the next decision.

First, a conversation

A direct conversation, no deck, no discovery theater. You bring the end goal. We pressure-test whether there's something real underneath it and whether we're the right two people to build it.

This is free because it has to be. We are two founders. We can only take work we believe in, which means we have to find out fast whether this is that. You leave knowing whether there's a problem worth solving here. We leave knowing whether we should be the ones to solve it. Sometimes the honest answer is no, and you get that answer in the room, not three proposals later.

If there's something there, we tell you what the next step would involve and what it would produce. Then it's your call.

Then, the scope

A fixed-fee engagement over a couple of weeks, with one deliverable: a written answer to what's worth building, what it takes, and what it's worth.

This is where most of the value gets decided. Not in the build. In the work of figuring out what the build should be. Most engagements discover the real opportunity is different from the stated one, and that difference is the whole game. Getting it wrong here is expensive in a way no amount of good engineering later can fix.

The fee is flat and it stands alone. It is not a credit against a build and it is not a sales call wearing a deliverable's clothes. The output is yours. If the answer is "build this," you have a scope precise enough to act on. If the answer is "don't build anything, fix this other thing first," you have that, and you have it in writing. Either way you could hand the document to anyone. That is the point. Work you would only value if it led to hiring us is not work worth paying for.

We will tell you if we don't think there's a build worth doing. That happens. It is cheaper for everyone to learn it now.

Then, if it's worth it, the build

Fixed price, scoped by the work that came before it. Not a guess. An estimate built on the answer we already reached together.

By the time we get here, the hard decision is already made. We know what we're building and why it's worth building. What's left is execution, and execution is what we do: orchestration, memory, self-hosted models where they earn it, the feedback loops that make the system improve on its own. We build it. Not a vendor we're reselling.

The build ends at self-sufficiency. The system runs without us and gets sharper the longer it runs. If it needs us forever, we built it wrong. Some teams want us to stay on for integration, updates, and maintenance as the system meets the rest of their operation. That's a separate arrangement, entered only if you want it. The default is that we finish and step back.

After that, only if you ask

We don't disappear, and we don't install ourselves as a permanent cost. After a build, some founders want a technical voice in the room when the next hard question comes up. We do that case by case, light, no standing retainer unless the work calls for one.

The relationship is the same one that ran through every gate before it. We're here when there's something worth deciding, and we're honest about when there isn't.