There is a version of AI consulting where the deliverable is a deck. Somebody who has never shipped a thing stands at the front of the room, walks you through a slide about your future, and hands over a roadmap that ends in a number. I have sat through plenty of those. I have also been the person shipping the product afterward, which is why those meetings make me itch.

So before I say anything about client work, here is my side of the ledger, with links you can click tonight. Parent Pep Talks is a daily pep talk for parents in the thick of it, with tips that adjust to the age and stage of their kids. Shut the Box is a game I built from scratch. Comfort & Vice is a shop. There is a fourth product, LGBTQueue, still on the workbench; it gets named here and nowhere else until it has earned its link. None of these live in a drawer. They are end-to-end consumer products, built collaboratively from the ground up, and running them is a standing education in what software costs after the launch party.

I have shipped at other scales too: I founded Section II, the first Benefit Corporation dedicated to LGBTQ+ women, and it reached a million monthly users. That one ran at a different scale. The subject here is smaller: what building real products teaches a person who then walks into your conference room.

Three things shipping teaches

1. A stranger breaks it in a day. A deck survives on plausibility because nothing in it can be pressed the wrong way. A product meets real people, and real people tap the thing twice, skip the instructions, and use it for a purpose you never planned. Every one of those surprises is information you were missing, and it shows up in hours instead of quarters.

2. Revenue changes the meeting. When your own product takes money from real people, every feature conversation gets a floor under it: what does this cost to run, who maintains it, what happens when it breaks on a Sunday. Opinions get cheaper fast. The questions get better.

3. Maintenance is the empathy. Anybody can be enthusiastic about a product in week one. Keeping one alive through month six teaches you what running quietly in the background demands of a person, from attention and small repairs to the willingness to look at a thing you made on a day when it misbehaves. A consultant who has worked that shift does not hand you a system and wave goodbye.

Why keep my own products going at all while the studio takes on clients? Because the alternative is advising from memory. The week I spend fixing my own game at midnight is the week that keeps my advice about your project honest about the part after the demo, the part where the thing has to survive contact with a real week.

What this means when you hire me

The case for a consultant who ships is simple: the habits transfer. A person who has carried a product through its first clumsy month does not confuse a demo with an outcome. A person whose products take real money does not get starry-eyed about a tool; she gets specific about the decision it moves.

That discipline is how the client work here runs. The case studies came out of it: a restaurant GM who now plans his week against one early forecast instead of four tabs, and a research nonprofit with a search method the team can rerun as the science moves. In both projects the starting point was the same: how the team already makes decisions, which is the 5% principle in practice.

So if you are hiring an AI product development consultant, ask the small question first: What did you ship, and who uses it? A deck can promise you a future. A shipped product has already survived somebody's Tuesday, and the person who built it learned things no slide can hold.

Ship first, advise second. That is the whole job description.