A driving school · Owner
Had dabbled in AI, never gone deep.
Runs teams of agents and dashboards he built himself.
Everyone knows AI could help. Almost nobody knows where to point it.
I teach you to build the thing that catches it, in your own tools, so you can build the next one yourself.
When a team wants one built rather than taught, that is a separate piece of work, and I do that too.
I learned to spot these the hard way. After losing my memory in 2020, every step cost me something I could feel, so I could not afford to drop the same thing twice. I still cannot unsee it.
They are your best insight into where AI could actually help you.
The answers are one size fits one. We find where yours are dropping, then I teach you to reverse engineer your own agents, inside the stack you already have.
You ask AI for a draft, then rewrite it. It guesses at what you meant, you correct it, and you do that again tomorrow. Nothing you taught it on Monday is there on Friday.
Your AI knows how you sound and how you work. It runs inside the tools you already use, picks up the thing you keep dropping, and gets better every time you give a correction back. Teams I work with go from never having used an agent to having agents working on their behalf, often in one session.
What gets built is different every time. One size fits one. The way we get there isn't, and it comes from how I rebuilt my own life. Zero to one first, then the how.
What ball are you dropping every day? The thing you put off, the thing that waits on you, the one that makes you think "ugh, I have to make time for that." That feeling is the signal. Most AI work fails here, pointed at something that was never the problem.
After my injury, my kids' memories were still in them. I had to learn how to get them out. Your team's know-how works the same way.
Ungrounded, AI gives everyone the average answer. We make yours explicit: what you sound like, what you never want to sound like, and how you want to be worked with.
Nothing came preloaded for me. I had to choose what good meant, on purpose.
Everything that lives in your head gets told to your AI, then we build. On your screen, in your tools, with you doing the typing.
I had to say out loud what I knew before I could rebuild any of it.
You give the fix back, because that's how it learns your version of good. What you learn stays in the documents and systems you own, so future you starts ahead.
You shouldn't have to lose your memory to learn this.
A driving school · Owner
Had dabbled in AI, never gone deep.
Runs teams of agents and dashboards he built himself.
An insurance adjuster · Solo practice
Had never opened Claude.
Three weeks later his agents correctly tagged 144 photos on a practice claim and filed them into his systems. He runs a team of agents he owns.
Jim · Mortgage brokerage
A working business with no formal website and nothing built for how AI finds a company.
One session later he built his own design system and had the site live inside 24 hours, written so search engines and AI assistants can read it. He now runs several agents that work on his behalf.
A staffing firm · Four owners
Using AI to draft an email, then rewriting it. Nobody had built an agent.
Each of them builds their own agents inside their system of record. One task that took 25 minutes now takes about 2.
I'm a systems engineer who has worked across a lot of platforms, usually as the person called when something breaks. I'm not tied to any of them.
Most people point AI at rewriting emails. I came to it with no picture of where it belongs, and no fixed idea of how a problem is supposed to be solved. It does not matter whether I have seen your industry or your system before. The way I unpack a problem and aim agents at it is the part that transfers, and it is what I teach you.
We work in the tools you already pay for. No new platform, no new vendor review, no new infrastructure. What changes is what you know how to do with what you have.
You do the building, so the capability stays when I leave. I teach your team to fish, as far as they want to go.
A gift · Free · No email required
Without grounding, AI defaults to the average of everyone else's good, which is why so much of it reads like a sea of sameness. Two interviews fix that. The first captures your voice: what you sound like, and what you never want to sound like. The second captures how AI should work with you, down to how you think and when to push or back off. If you needed an evening gown you might buy one off the rack, but you would have it tailored. This is the tailoring, and it is what closes the gap between an answer that is 85% right and one you can actually use.
Get the two promptsUse them on your own, nothing required. Where they live in your stack and how to put them to work is what we do together. See what else is on the Shelf
Whiterock Road is Jenny Cotie Kangas's Builder Coach practice. She teaches people to build their own AI agents and automations inside the tools they already use, starting with the work they keep dropping. She is not building it for you and handing it over. You build it, with her next to you, so the capability stays with you.
Someone who builds with you and teaches you to build, so the capability stays after the work ends. In Jenny's words: "I'm teaching how to build. I'm not doing it for you. I'm meeting you where you're at and helping you go from zero to one."
Start with the ball you keep dropping, not with a tool. The task you put off, the thing that waits on one person, the work that never quite gets done. Those are your best insight into where AI would actually help. Picking the right one is the first thing we do together.
Common first builds: a front door agent, which is one place your work starts and which knows how you work and what your systems hold; a daily brief that pulls together what matters before you open your inbox; expense tracking that happens without you remembering; or an agent that knows your CRM and does the configuration and optimization work that keeps waiting. The answer is one size fits one, so it depends on what you keep dropping.
No. The work happens in the stack you already pay for, so there is no new platform, no new vendor review and no new infrastructure cost. What changes is what you know how to do with what you already have.
Book a 30-minute intro, Tuesdays and Thursdays, or email jenny@whiterockroad.org. Nothing is required beforehand. If you want to arrive further along, the two free prompts at whiterockroad.org/two-prompts teach your AI what you sound like and how to work with you.
This work is fast. The ball you keep dropping, handled by something you built. Set future you and your future team up for success.