Who this is for (and who it isn't)

At some point in the last year, somebody — a vendor at a trade show, a nephew, a man on LinkedIn with suspiciously white teeth — told you that AI is about to change construction forever. You nodded, because nodding is easier than asking a follow-up question at 4:30 on a Thursday, and then you went back to chasing a draw request through three spreadsheets and an email chain with eleven people on it.

This book is for the version of you that wanted to ask the follow-up question.

Who this is for

You own or help run a small-to-mid-size construction company. Maybe you build homes. Maybe you're a GC, a sub, a remodeler, a small developer with a couple of projects going and a couple more you're trying to pencil. Whatever the letterhead says, the actual job description is the same: you are the estimator's backstop, the bookkeeper's double check, the IT department, and the person who knows why the printer is unplugged.

Your data lives in more places than your tools do. Estimates in one spreadsheet, job costs in another, contacts in your phone, schedules in your head, and the real story of every project scattered across text threads. When the lender asks a simple question, producing the answer takes a day plus and makes you look worse than you are. You've suspected for a while that AI ought to help with this — and every time you've tried to find out how, you've gotten either a sermon about the future of work or a listicle of forty apps, thirty of which will be gone by Christmas.

What you actually want to know is where the real work is. What to do first. What it costs. What breaks. Whether a company your size can do this without a CIO, a consultant on retainer, and a second mortgage.

That's this book.

A word on who's writing it. I'm Donald Speedie. I build homes in California's Central Valley through Nimble Development, with my partner Manny, a general contractor, running the field. I've been building for 28 years, and along the way I started building software for builders too — that company is FluidCM. And about a year ago I made a bet that sounds ridiculous when you say it out loud: instead of hiring for the back office, I would teach AI to run it. Not a chatbot bolted onto the side. The actual operating system of the company — books, forecasts, deal screening, scheduling, follow-ups, the works.

This book is the complete lift of what we built, what it cost us to learn, and how you could do the same thing without paying our tuition.

Who this isn't for

Fair warning, because your time is the one thing nobody can refund.

If you're here for a tool roundup — "Top 40 AI Apps for Contractors" — this isn't that. There's a full stack list in Appendix A, every tool named with what it does and what it costs, but I put it in the appendix on purpose. The tools are the least durable part of the whole system. Reading this book for the tool picks is like buying a house for the paint.

If you're here for an AI hype fix, same door, other direction. I'm not going to tell you AI will triple your margins by Friday, replace your estimator, or "revolutionize" anything. I run on this stuff every day and I'll show you exactly what it does — and a whole section on where it fell on its face. If you want breathless, the internet has you covered.

And if you're planning to skip the boring chapters — the ones about cleaning data, naming sources of truth, and writing down rules — I'd gently suggest that's how you ended up with three spreadsheets and an eleven-person email chain in the first place. The boring chapters are the system. Everything impressive sits on top of them.

The promise

Here's what you'll have when you finish:

First, a working picture of what "AI-first" actually means at our scale — not the conference-keynote version, the Tuesday-morning version. Part I covers the foundation that has to exist before any AI earns its keep, and Part II is the full tour of the system running my company today: what each piece does, what it replaced, and what it costs to keep running. Real numbers where I can share them — like the enterprise-grade accounting setup we stood up for about $5,000, or the $25-a-day leash the whole system runs on. Where I can't share a number, because it touches a specific deal or a client, I'll say so plainly rather than make something up.

Second, the scars. Part III is every lesson that cost us something — the automated systems that quietly rotted, the alerts I trained myself to ignore, the day I discovered 548 findings piled up that nobody, including me, had read. Cheaper to read about than to live through.

Third, a blueprint. Part IV turns all of it into a phased plan you can actually follow: a readiness scorecard, the order of operations, honest costs by phase, and a clear-eyed look at what to build, what to buy, and when it's worth having a guide. That last part is where I come in, if you want the help — but the book stands on its own either way.

How to read it

Top to bottom, like a book — that's how it's built, and each chapter hands off to the next. But nobody's checking. If you're skeptical, start with Part III and read the failures first; I would. If you're impatient, jump to Part IV and work backward. If you just want to see the machine, Part II is the showroom. Chapters cross-reference each other by number, so you can wander without getting lost.