Yesterday I told you why The AI Executive System exists.
Today I want to show you exactly how we built it — and what went wrong along the way.
Because between “here’s a cool idea” and “here’s a program that works,” there’s a stretch of hard work that most companies don’t even bother to do, and the rest rarely talk about publicly.
That’s today’s topic — what 200+ of you told us you needed, what 100+ students actually struggled with in the first cohort, and how we solved those challenges for current and future students.
Then tomorrow, enrollment opens so you can decide if now is the right time for you.
Let’s start at the beginning.
—
Part 1: What 200+ of you told us
Before we built anything, we ran a deep survey with our community. We worked through every response by hand and built a thorough dataset from the answers to inform our “V1” of The AI Executive System (AIEX).
Chart: Roles, All High Performance Professionals
Four things came back loud enough that we designed the program around them.
1. You’re not stuck on how to do great work. You’re stuck on how much work there is to do.
The pain wasn’t knowledge. It was capacity and throughput. The sentiment, over and over, was some version of: I know exactly what to do. I just can’t get it all done without something breaking or me getting burned out.
2. You were afraid AI would make your work worse.
You’re already delivering at a high level — that’s how you got to where you are today, and not something you are willing to give up under any circumstances. We are not “outsourcing ourselves” to low-quality automations.
That means the real question is not “should I use AI more” or “can I do this faster.” It’s: How do I move faster while also maintaining or increasing the quality of my output.
3. Personal life was taking a big hit.
Nights. Weekends. Recovery time. The feeling of “I can’t keep doing it this way for another three years.”
Chart: Problem Mix, All High Performance Professionals
4. Three years from now, you don’t want to just be more efficient. You want to be better.
Short term, many of the answers were about “survival” at work. But stretch the horizon out and the tone changed — professional mastery, leadership, stewardship, being a better partner and parent and neighbor.
Not make me faster. Free up the capacity so I can be the best version of myself.
That last finding is the reason AIEX is a six-month program and not a weekend bootcamp. You told us you wanted structure, guided implementation, and room to practice without it becoming a second job.
CHART: Support Preference
—
Part 2: So that’s what we built
Three phases, mapped directly onto what you said:
Stabilize — reclaim the week first. The AI agents that do serious work: inbox, calendar, scheduling, expenses, the project planner.
Systemize — this is how to produce better and faster results at the same time. You build a library of your own expertise — your documents, your past work, your voice — and the assistants work from it. So the output doesn’t sound like a chatbot. It sounds like you, with infinite stamina.
Scale — the long-horizon layer. An advisory council, a chief of staff that routes work across your specialists, and a home manager for the personal side.
That’s the plan. Here’s what it looks like in real jobs for real students.
Jimmy A.
Jimmy and the weekly report
Jimmy is an executive running the smallest team in a 1,000-person company — and also handling the biggest footprint of any team. His weekly management report used to burn hours and hours out of every single week.
“And obviously my hourly cost is a lot of money, and it’s like — is that worth it for the company?”
So he trained his AIEX agent on ten years of his own reports. Then he started dropping little chunks in as he went through the day — here’s what I did, here’s a meeting transcript.
“Now it takes me about 30 minutes for the entire week to produce the whole report and send it to my management team.”
I asked him the obvious question: is it better than what you were doing before?
“Oh, 100%. It’s my content, but it’s all generated by AI at the end of the week.”
His leadership noticed, even before he told them anything:
“Our CEO and our president have both been like, I don’t know how you do it. I don’t know how you have the time to produce such a big report… Everyone’s like, ‘Dude, you have the best reporting of any group in the entire company.’”
Then he taught the whole thing to his team.
“Everyone’s like, ‘You guys are, like, the most high-performing team’ — and we’re also the smallest team in the company managing the biggest footprint.”
That’s finding #2, answered. Faster and better, in his own voice, on his own material.
And on what he does with the hours:
“I literally can do eight hours of work in about three hours, and nobody believes me until they see my output. And I’m like, ‘Guess what? I worked half days this whole week.’”
Deb S.
Deb and the org chart
Deb runs a consultancy outside Philadelphia (plus real estate and non-profit work on the side). Six months ago it was her and a VA.
She’s now built 28 agents, and they’re not just a pile of tools. They’re a coherent team.
“All the agents are connected to the chief of staff and the house manager. And then those two agents are connected to the CEO, and the CEO watches over the entire umbrella. And then all the agents that need to be connected are also connected to each other.”
Calendar talks to email. News feeds her learning agent. The learning agent feeds a journal she keeps for anything that comes up “in the entire process of life, running a business and family life.”
Half of those 28 run her business. The other half run her household.
And this is finding #3 and finding #4, in one sentence:
“Before, I would in the middle of the night have to do my learning… because the entire day, weeks, months are focused on trying to figure out the business.”
She now does that learning during the day. Cybersecurity, AI ethics, a new language. And she’s using the rest of the time to help family members set up their own systems and start their own businesses.
How she got from “me and a VA” to a 28-agent org chart in six months is a story in itself — and I’m going to tell you the whole thing later this week. Keep an eye out for it.
—
Part 3: Three things we didn’t see coming
We ran this with 100+ founding members for six months. One of them described the experience to me on camera as “a little bit of building the plane while you’re flying it.”
She’s not wrong — that’s how we intentionally run every founding cohort, because we’re not just posting a video and forgetting it. We’re creating a system that you’ll be able to use for years to come. Real experience, including addressing real problems, is essential.
Here are the challenges we hit as we built AIEX.
1. We were surprised by how bad a lot of off-the-shelf agent software is.
I have been developing software for 20+ years, so I know there’s a lot of junk out there. But I was really surprised how half-baked so many of the popular tools were (and still are today).
I think there are two big reasons for this — first, common automation tools like Zapier and Make pre-date AI, so they are basically being retrofitted for the new era.
Second, nobody really knows what the business model is for “agent-builder tools.” Are they going to cost $20/mo? Are they going to get acquired by Anthropic? Is Silicon Valley even going to exist in 5 years?
The result of that uncertainty is that many agent-related software products that seem exciting eventually get abandoned or go stale. You end up with sort of a sad graveyard of promising half-built tools that don’t really work for most people.
We went in expecting to plug into existing tools and teach on top of them. So we tried a bunch. And we ended up scrapping several tech recommendations because we could not get students to a place where the thing actually worked.
The question I heard more than any other in those early months:
“Is everything this buggy?”
And I knew the answer was no — it doesn’t have to be that way. But there is a ton of very popular software out there that is, honestly, very bad at building real, working agents and assistants. Slick demo, great landing page, then three hours of banging your head against a wall until you give up.
(If you’ve ever tried wading through all the AI tool options yourself, I suspect you figured this out a while ago.)
So one of the ways we answered it was to build our own.
IWAI Studio is a free, portable, flexible app we made specifically to give people a Fast Start — talking to your first AI coworker in about half an hour instead of losing a weekend to setup. Some students use it forever. Others use it to learn the concepts, then take what they’ve built and jump to other tools, which is entirely the point: everything is exportable.
It did not exist in January. It exists because of what the first 100 students ran into.
2. The ground moved underneath us — even faster than we planned for.
Back in her first weeks in the program, Deb told me the exact thing that almost stopped her from joining:
“What if the technology moves faster than the training?”
Deb knows her stuff, and this is a great question.
During the first few months of our first cohort, the entire arc of OpenClaw — released, went viral, everybody lost their minds, changed names twice, turned out to be a nothingburger — happened from start to finish. That’s the pace we were keeping up with.
The industry is not going to move any slower in the next six months. And so, what we created in AIEX is not “here’s a tool, memorize it.” It’s a set of tech tracks that are flexible and logical, where new things slot in as they arrive: a Fast Start path, an integrated-suite path, and a fully self-hosted path for the highest level of privacy and control.
Jessica B.
Which brings me to Jessica.
Jessica is a clinical neuropsychologist in Boston. She also runs a data platform company in Portugal and a nonprofit doing early relational health work in Mexico.
“My clinical data can’t touch the internet, like, really in any way.”
Because of healthcare privacy rules, using cloud AI like ChatGPT and Claude is a non-option. On top of that, she started out with nearly zero tech knowledge:
“My relationship to tech in general, and the whole AI revolution, was fear and trepidation… I had no technical knowledge before, really, at all. I used email and I used the internet to search things, but that’s about it.”
She was tracking time for her Mexico team with a pencil and paper. Not a spreadsheet. Paper.
Here’s where she is now:
“What AIEX taught me is that that [private] server can be a Mac Mini — and I can actually create the stuff that needs to be on it without becoming a programmer.”
That’s the kind of transformation we’ve spent the last 6 months designing for every AIEX student in all our different tech tracks.
Unlike a lot of YouTubers and so-called course creators, we are not selling you a tutorial for one piece of software that may or may not exist next year. We’re teaching you how to build these systems several different ways, so that whatever models, tools, or buzzwords show up in 2027, you already know how to absorb them.
3. Coaching had to work completely differently here than in our other programs.
In our consultancy program, small-group coaching is a crowd favorite. People running AI consultancies tend to hit the same set of problems eventually — pricing, scoping, the awkward client conversation — so sitting in a small group and hearing someone else’s question is often more valuable than asking your own.
AIEX is not like that. The live training calls have been great for building community, and they’re continuing as AIEX Demo Days. But for our Coaching tier students, the more pressing need turned out to be something a standing weekly call is structurally bad at: debugging your specific build.
Jessica put her finger on exactly why:
“The group sessions were good, but when you’re working and doing 100 different things, it’s hard to show up with the most pertinent questions in the moment.”
And more bluntly:
“The one-on-one coaching has been critical, just because I am so non-technical that often I need help just finding the button.”
Nobody wants to say “I can’t find the button” on a call with 20 people. But it’s the difference between finishing and quitting. Here’s how Jessica described her experience after 1:1 calls with Nyasha, one of the tech coaches here at IWAI:
“It wouldn’t have happened without her. I mean, it just wouldn’t have. I would’ve given up.”
The other problem is timing. When your build breaks on a Wednesday, you need to be unblocked quickly — not on Monday when the next call comes around.
So we introduced two things that didn’t exist in January.
AIEX Fast Passes. Coaching-tier students book a session, screenshare, and sit down with our actual coaching team — real software developers — to work through their exact build. Not just “here’s how to think about it.” Here’s your bug, fixed, in your environment.
Jessica again, on why that format specifically works:
“I don’t have that kind of time. But I do have an hour to spend with Nyasha and fix the problem that I have — and that makes me wanna do it again when I have another problem.”
And then something entirely new: the Executive Concierge tier.
Jessica, Jimmy and Deb are proof that you can build this yourself with no prior coding background.
But building takes time, and some people are going to get more out of having a pro build the first \~10 agents for them — and then expanding and iterating from a system that’s already running. So that’s exactly what we now do in the Executive Concierge tier: our team works with you across a structured engagement to build them, launch them in your environment, and then teach you how to run and extend the whole thing yourself.
It’s the most hands-on level of service we’ve ever offered, and spots are limited by how much team time each engagement takes.
I’m especially excited about this because it merges my 20+ years of client service work with everything we’ve built at IWAI.
—
What one of them said about all of it
I asked Jessica how she’d describe our company after six months of the plane-building-in-flight experience.
“Knowledgeable, friendly, real… willing to take risks.”
“You figured it out. You didn’t tell us we were crazy. You didn’t say, ‘Stop asking questions.’ You figured it out.”
That’s the standard. Not “we got it perfect on the first try.” Nobody does, especially not in such a fast-moving and fragmented industry. We figured it out, in public, with the people who trusted us early.
—
Tomorrow
Enrollment opens tomorrow, exclusively to IWAI readers.
I’ll walk you through the whole thing: the three phases, the 20+ assistants you’ll build, the three tech tracks, the new coaching options including Fast Passes and Executive Concierge, and exactly what’s included at each level. You’ll be able to look at all of it and make a fully informed decision.
No pitch today. This was just the account of how this program got made.
Two hundred of you told us what you needed. A hundred of you spent six months finding every rough edge with us. Tomorrow we share the result.