Questions people ask before they call
Do we need developers on our side?
No. The assessment, workspace setup, skills library and training are done with the people who will use the tools, whatever their job is. A custom product build needs one person on your side who can make decisions and say yes or no to what I show them. Software teams are welcome too; the method and the integrations are built for them.
We already use ChatGPT. What would actually change?
The output. Most teams use AI as a blank chat window, so every result depends on who typed the prompt. Setting it up properly means a shared library of tested skills, the model connected to your actual documents and systems instead of the open web, and a scorecard that says whether the output is good. Same tools, different results.
Where does our data go, and do you train on it?
Your data goes to the AI provider you choose, under its business terms, and nowhere else. I don't train models on your data, I don't share it, and I don't keep it after the engagement unless you ask me to. Before anything connects to a live system we agree in writing which data it may read and which it may never touch. The detail is in the privacy policy under "AI and your data".
How long does it take?
The first call is thirty minutes. The assessment takes one to two weeks. Setup and skills packages usually take two to four weeks. A custom product is scoped per project and the plan says how long before work starts.
What does it cost?
The first call is free. The AI Readiness Assessment is $2,500 flat and is credited in full to any build within 90 days. The setup packages are fixed fees quoted after the assessment, because the scope depends on what it finds. Custom builds start at $12,000. The partner retainer is $6,000 to $10,000 a month. Hourly consulting is $200 an hour. Every engagement is quoted in writing and invoiced.
What don't you do?
I don't resell AI licenses or hardware. I don't build systems that act on customers, money or records without a person approving the step that matters. I don't promise a number before the assessment. And I don't take on work where I'm not the one doing it.
Who owns what gets built?
You do. Skills, automations, connectors and custom products built for you are assigned to you as stated in the proposal, and they're deployed to accounts you control. I keep the right to reuse general methods and tooling that aren't specific to your business. If you need it worded differently, we put that in the proposal.
Which AI tools do you work with?
Whichever your business already pays for or is comfortable with. I build with Claude most, and I work with ChatGPT and Copilot setups too. The connectors use the Model Context Protocol, an open standard, so they aren't tied to one provider.
How is this safe? What stops the AI from doing something wrong?
Structure, not trust. Anything that spends money, contacts a person, or changes a record stops for a human approval, and the dangerous actions are built so the AI has no way to do them on its own. Every system on the projects page works that way. And nothing deploys itself: every change to a running system is a person's action.
What if it turns out AI isn't worth it for us?
Then I'll say so on the first call, and it costs you thirty minutes. If the assessment finds the payoff is small, the report says that, and you keep the one working result it produced.