Start with the problem.

Maybe a process takes too much work. Maybe you already bought the tools and nobody’s using them. Or you have an idea that AI might make possible, but you’re not sure where to start.

I like looking at those questions with the people who understand the work. What’s causing trouble? What have you tried? Who will have to run whatever we build? Sometimes we need to build something. Sometimes the useful answer is simpler than the one already on the table.

Tell me what you’re trying to accomplish, what you already know, and where you’re stuck. We can work out whether I’m the right person to help.

I could take an independent look at a proposed design or an existing system. We’d agree on what to review and what you need from it before starting.

I’d want to work through questions like these:

• Does this solve the problem you actually have?
• What does it depend on, and is any of it more complicated than it needs to be?
• Who can access it, how would you recover it, and who will look after it?
• Where could AI or automation help, and how would we check the result?
• What other options are worth considering?

You should come away understanding the choices and knowing what to do next. The cases here give you a sense of how I’d approach it.