Pre-AI consulting · The decision before the build

What should AI never do in your business?

It's a strange question, and almost nobody is asking it. Everyone wants to tell you what AI can do. I help you decide what it's for — what to build, what to wait on, what to stop, and what it should never touch. Then I hand you the map and go home.

1-855-DREWSAI

A real person answers. Eventually it's me.

Scene — decision island A small human desk at the center of a paper-craft island. Three roads lead out: one to a green check, one to a violet pause, one to a red stop. A slightly imperfect red rope wraps the island, and a goofy over-engineered machine waits politely on the other side of it.

The part everyone skips

You were told to make AI real. Nobody decided what it's for.

The board asked, the budget moved, and the vendors started calling — and underneath all that motion, one small thing is missing: a decision about what the AI is actually for. That's why the pilots stall. It was never a tool problem, and a better model won't fix it. It's a decision problem, and the decision is the part I do.

I'm not an agency and I don't build anything. I come in before the build, learn how the work really runs, and write down the decision everyone has been working without.

Futurism in motion

The future is coming either way. You decide how to walk into it.

You don't have to chase it, and you can't outrun it. What you can do is know your four answers — and a company that knows them moves faster than the one that bought the most software. It isn't close.

Buildthe few things worth doing first, and why those first.

Waitwhat can wait, and what you're actually waiting for.

Stopwhat to stop now, before it quietly costs real money.

Neverthe things AI should not touch in your business at all.

Every answer I give names at least one thing to stop, in writing — because a decision that only says "go" is a sales pitch wearing a nicer shirt.

The line nobody prices

The most valuable line in your business is the one AI should never cross.

Your client's privileged files. A doctor's judgment. The hiring call. The relationship your customers actually pay you for. I draw that line in writing, clearly enough that your board and your regulator can both see exactly where AI stops.

Everyone will sell you what AI can do. I put in writing what it must never do here — and that one page tends to outlive every tool you'll buy this decade.

Scene — the never-line A human desk on the left, the over-engineered contraption on the right, and one confident, hand-inked red line drawn across the ground between them. The line is the hero of the picture.

The deal

Here's the deal — and it's my favorite part of the business.

I don't build it. When the decision is made, I'm done — I'm not quietly auditioning for the build job.
I won't become a monthly bill. No retainer, no seat that renews while you forget it's there. If you need me again, you call on purpose.
Every answer names something to stop. In writing, every time.
I draw the never-line. Also in writing, with your examples, not abstractions.
My prices are on the wall. No back room, no "request a quote."
You're hiring my judgment, not a platform. My name is on the work and on the phone.

Each one of those closes a door other firms leave open on purpose. That's not me being noble — it's the design. Advice you can trust has to come from someone who makes the same money either way.

I know how this sounds

A consultant who builds nothing, sells nothing afterward, and will happily tell you to do nothing.

It sounds like a fool's errand, and I run it cheerfully. Sometimes the honest answer really is "not yet," and you'll get it with a straight face and a one-page brief. Almost everyone else in this market makes their real money on what comes next — the build, the tools, the seat that renews. I set the business up so the only thing I sell is the decision, which means it's the one thing I'll never shade.

Scene — guide walkaway A plain paper-craft traveler — a map in hand, no toolbox in the pack — hands over a rolled map and walks off down the road, waving. One of the roads behind him honestly ends at a small red marker.

Every price, in public

Four ways in. Every price is on the wall.

The Question$8.55 One real question, one straight answer, fifteen minutes. The Room$2,500 A public AI resource for a community — your name on it, no agenda in it. The Sprint$8,500 One workflow, one honest week, a decision you can act on Monday. The Crash Course$85,000 The whole company, two weeks, the full map.

Yes, the numbers rhyme with the phone number. 855 · $8.55 · $8,500 · $85,000. It makes them easy to remember, which is the point of putting prices on the wall.

Who actually answers the phone

You're hiring a person, not a platform.

I'm Drew. For years I did the unglamorous version of this work — analytics, product, applied research — figuring out what a business should actually do with its data, and then helping it do exactly that and nothing more. I started Drew's AI because I kept watching good companies get sold an AI build before anyone decided what the AI was for.

One active client, under NDA. I won't show you logos I can't name or quotes I can't stand behind — here's what I can show you, and exactly what I can't.

It's a funny way to run a business. It's also the most honest one I know.

Call me with the question you came here with. If you don't need me yet, that's the answer you'll get — and it'll be the best $8.55 you spend this year.

1-855-DREWSAI

Curious what happens when you call? Here's the whole thing, step by step.