The first 30 minutes of your AI Readiness Audit

Most "AI consultants" open with what model should we use? That's the wrong question. Here's what I actually ask in the first 30 minutes.

The 5 questions I ask, in order

1. Where do you lose time today? Not where it's annoying — where it's expensive. Ops, support, sales follow-up, content, reporting.
2. Which of those would be fine if a smart junior did it? That's the AI fit signal. Creative judgment, client relationships, and negotiation stay human.
3. What data do you already have, and where does it live? If the answer is "everywhere and nowhere," we start with an integration layer (usually n8n) before touching a model.
4. What does "done" look like in 30 days? Hours saved per week, dollars recovered, leads contacted. If you can't measure it, I won't build it.
5. What have you already tried that didn't work? Reveals your pain tolerance, your stack, and the shape of a fix that'll actually stick.

The 3 decisions I make

• Build / buy / kill. For each workflow: custom automation, off-the-shelf tool, or "don't do it yet."
• Reasoning vs. plumbing. Where Claude lives (read, summarize, classify, write) vs. where n8n lives (trigger, route, schedule).
• Sequence. Which flow we ship first. Almost always the one with the highest margin × lowest integration cost.

The 1 deliverable

A short written scorecard with:

If the scorecard isn't useful, the audit's free. If it is, you have a map to the next 90 days — with or without me.