AI Business Build Week · Free 5-day live build

What changes in five days

You don’t leave with notes. You leave with files you’ll open again next week.

Each day’s file is the next day’s input. Day 1’s Offer Card is the first file Day 2’s Business Brain reads. The opportunity you rank first on Day 1 becomes the agent you spec on Day 2 and the automation you test on Day 4. Nothing is a standalone exercise.

Two kinds of people. One week. Pick honestly. The rule I use on Day 0: choose OPERATE if you’ve had paying customers in the last 90 days and you repeat work you could map. Choose LAUNCH if you don’t have customers yet, or you want a new offer that’s separate from your current business. If both fit, pick by the outcome you want after Day 5. You’ll see both tracks built live every day either way.

LAUNCH

I’m starting from zero

By Day 5 you’ll have built v1 of everything a solo AI business starts with:

Day 1Offer Card
Day 2Business Brain v1 + 4-role AI team
Day 3Offer Page Kit v1: live page + lead magnet
Day 4First-Sale System v1 + prospect list
Day 530-Day Launch Plan + Solo OS v1

Before Day 1: An idea, a few tools you’ve signed up for, and no clear first customer.
After Day 5: One chosen offer with evidence behind it, an AI team that knows your business, a page at a public URL, a path from stranger to booked call or paid test, and a 30-day plan with a weekly scoreboard.

LAUNCH is for you if

You have a skill people already ask you for help with, and you want to turn it into a first offer you can test this month.
You’ve collected AI tools and tutorials but haven’t put one thing in front of a buyer.

See how each day runs

OPERATE

I already run a business

By Day 5 you’ll have built v1 of an AI operating layer for your business:

Day 1Ops Map + AI Opportunity Scorecard
Day 2Company Context Pack + first agent spec
Day 3Content Engine v1
Day 4Automation Pair: speed-to-lead in test mode + back-office spec
Day 5AI Operating Layer v1: dashboard, policy, 90-day roadmap

Before Day 1: AI is “something we should look at”. Knowledge lives in your head and your inbox. Nobody can say where AI would actually pay off.
After Day 5: Your recurring workflows listed and the biggest ones scored, a context pack your AI reads before it answers, a content system sized to the time you really have, one revenue automation in test mode plus the spec for a second, each with an ROI sheet that keeps hours and cash apart, and a 90-day roadmap with kill rules.

OPERATE is for you if

You run a service business, a studio, a practice or a small team, and you decide which tools it uses.
You want to know where AI pays off in your business before you spend on it, and where it doesn’t.

See how each day runs

It’s not for you if

You want a promise of income. I won’t make one.
You want to watch rather than build. The value is in the five things you make.
You can’t find about two hours a day for five days: the live hour plus the homework path, and on OPERATE a short prep block on Days 1, 3 and 4.

Every deliverable is a v1. It’s meant to be used and improved, not framed. You build the assets; what they earn is up to your market and your work.

How the week runs

Day 0 to Day 6. One rhythm, every day.

Before Day 1, you set up your Build Desk: one project folder, one tracking sheet, one brief your AI reads first, and only the accounts this week needs, free tiers first. Then five live days, each with one principle and one deliverable per track. Submit all five and you’re a Builder, which gets you a seat at the Day 6 Implementation Clinic.

Day 0 · Before Day 1

Set up the Build Desk: one folder, one tracker, one brief your AI reads first.

My rule from agency work: no build starts until access, folder, brief and baseline are in place. AI makes that rule stricter, not looser, because an assistant with no context fills the gaps with plausible guesses. Day 0 gives every later day the same starting point.

00-START-HERE.md

# 00-START-HERE
## Owner and track
## What I’m building (one sentence)
## Who it’s for
## Known facts
## Assumptions
## Unknowns
## Constraints
## Working rules for AI
## Decision log

The headings of the one-page brief your AI reads first, from the Day 0 pack.

Both tracksbuild

Build Desk

Homework path (about 40 min)

Your track chosen with a written reason, an account plan (what each tool costs, what you already have, what’s blocked where you live), one build-week/ folder, a Build Week tracker sheet, and 00-START-HERE.md, a one-page brief that separates facts, assumptions and unknowns.

LAUNCHadds

A raw idea inventory for Day 1.

OPERATEadds

A tool inventory and a last-month numbers sheet, each figure marked measured or estimated.

Day 1 · Principle: AI multiplies a good offer; it doesn’t create one

Decide where the money is, before you build anything.

The expensive mistakes in the projects I’ve managed usually started at the brief, not the build: the wrong problem, a buyer nobody had pinned down, a price nobody had tested. Day 1 forces the decision every later day depends on, with evidence instead of enthusiasm.

LAUNCH · Offer Card

OFFER CARD v1
Tag every line [FACT], [ASSUMPTION] or [GAP].

Offer name (plain words):
Who exactly (observable traits + where I find 10 of them):
Their problem, in their words (quote + evidence ID):
Promise (outcome I control + timeframe + boundary):
What they receive (3–5 concrete deliverables):
Price hypothesis (number + test range + reasoning):
Kill criteria (what makes me drop or reshape this):
Validation plan: 10 conversations by [date + 7] (initials or V1–V10 only; real names stay in your private notes)

A real page from Day 1This is what you fill in after Day 1’s live hour.These are eight of the card’s fields, copied from the LAUNCH Offer Card in the Day 1 pack. It isn’t a pitch. It’s a hypothesis with kill criteria attached, which is the difference between an offer you test and an offer you hope for.Open the full Offer Lab

LAUNCHbuilds

Offer Card

Homework path (≤40 min)

Your Offer Card, offer-card.md. Three rough ideas scored against real buyer language (reachability counts double), one chosen offer with a promise, a price hypothesis and kill criteria, and a script for 10 validation conversations.

Also in the pack

The 15-minute evidence-hunt guide, a thin-evidence mode for when you find fewer than eight buyer quotes, and a pointer to the full seven-prompt Offer Lab for a deeper rebuild after your first five conversations.

OPERATEbuilds

Ops Map + AI Opportunity Scorecard

Homework path (≤40 min, after about 30 minutes of prep that needs no AI: your numbers and recurring tasks gathered)

Your Ops Map + AI Opportunity Scorecard, one sheet. Your recurring workflows listed with owner, volume and hours, the top 3 AI opportunities ranked by value × feasibility × risk, and a This-Week Pilot Card for #1 that you can run by hand on real work.

Also in the pack, as a 20-minute stretch

A sceptic pass on opportunities #2 and #3, a step-by-step map of your top 5 workflows, and the “no AI needed” list.

Day 2 · Principle: context beats prompts

Stop starting every chat as a stranger.

A clever prompt can’t make up for an AI that doesn’t know your business. Today you write down what only you know: who you serve, what you sell, how you sound, what you never claim. Then you give AI clear jobs and test it on real tasks before you trust it, the way I’d onboard a new developer.

business-brain.md Fictional worked example

[FACT] Independent wedding photographers, solo or with one assistant.
[ASSUMPTION] They will pay for setup but not for a monthly tool.
[GAP] Q: “What if the date is already booked?” A: no answer yet; I’ll ask on my validation calls.

Every line of the Business Brain is tagged. A [GAP] is never filled by the AI; it asks you instead.

LAUNCHbuilds

Business Brain v1 + 4-role AI team

Homework path (≤40 min)

Business Brain v1 + a 4-role AI team. business-brain.md with every line tagged FACT, ASSUMPTION or GAP, four role cards (Researcher, Writer, Builder, Assistant), and the Writer and Researcher tested on real tasks.

Also in the pack

The HQ project setup in Claude or ChatGPT, a before/after pair that shows the difference the Brain makes, and the Builder and Assistant tests, scheduled for Days 3 and 4.

OPERATEbuilds

Company Context Pack + first agent spec

Homework path (≤40 min)

The Company Context Pack + first agent spec. One document under 1,500 words covering offer, customers, journey, voice, policies and data rules; your Day 1 top opportunity written as an agent job description with human checkpoints; and the three tests that matter most run and scored, including one that must escalate and one adversarial.

Also in the pack, as stretch

The two remaining test cases and a draft-only Agent Project you use on real work for the rest of the week.

Day 0 takes about 40 minutes, any time before Day 1. Each day card below splits the homework path (about 40 minutes, one must-have file) from what else is in the pack, and names any prep the day needs.

The live hour

Same 60 minutes every day. Ten minutes of principle, the rest is real work.

I don’t lecture for an hour. The principle takes ten minutes. In the 25-minute live build I work on two fictional businesses side by side, one per track, both labelled as fictional, and take each deliverable through its key decisions and its hardest step, so you see the choices and the checks, not just the result. The Day Pack covers the rest. Then the hot seat puts participants’ real work on screen.

Minutes 00–05
Wins wall 5 min
A few of yesterday’s submissions on screen, only with permission
Minutes 05–15
Principle 10 min
The principle of the day, shared by both tracks
Minutes 15–40
Live build 25 min
The key decisions and hardest step of the LAUNCH and OPERATE deliverables, side by side
Minutes 40–52
Hot seat 12 min
Participants’ deliverables reviewed live, one per track where time allows
Minutes 52–60
Homework 8 min
Tonight’s homework path and what tomorrow builds on

Every session opens with a recording notice. You choose whether your work appears on the wins wall.

What you get

Every day ships as a Day Pack you can use without me.

The live hour shows you how. The Day Pack is what you build with. Each one is written to the same standard I hold my own prompts to: it asks before it guesses, it separates facts from assumptions, it forces structured output the next step can use, and it tells you when not to use AI.

Open now · freeDay 0 pack: set up your Build Desk

1Day Pack
Decide where the money is
LAUNCH
Offer Card
OPERATE
Ops Map + AI Opportunity Scorecard
Ships on Day 1
2Day Pack
Build your AI team
LAUNCH
Business Brain v1 + 4-role AI team
OPERATE
Company Context Pack + first agent spec
Ships on Day 2
3Day Pack
Make the thing people see
LAUNCH
Offer Page Kit v1: live page + lead magnet
OPERATE
Content Engine v1
Ships on Day 3
4Day Pack
Automate the money moments
LAUNCH
First-Sale System v1 + prospect list
OPERATE
Automation Pair: speed-to-lead in test mode + back-office spec
Ships on Day 4
5Day Pack
Run it like an operator
LAUNCH
30-Day Launch Plan + Solo OS v1
OPERATE
AI Operating Layer v1: dashboard, policy, 90-day roadmap
Ships on Day 5

A Day Pack per day, per track: a worksheet with a filled example (fictional and labelled), the prompt chain with a self-check on each prompt, a quality checklist, and a homework path that marks what’s must-have and what’s stretch.
The prompts: 3 to 7 per day, each handing off to the next, with the inputs to paste, the expected output and a “reject the output if” line.
Claude skills and automation recipes where a day involves a build, including the Business Brain and Content Engine skills and a Speed-to-Lead recipe that works in Make or n8n.
Tools and costs listed before you start each day: what’s free, where a paid plan helps, and what may be blocked in your country.
Builder status after five submissions which gets you into the Day 6 Implementation Clinic.

No dollar values on any of it. Judge it by opening the free day tools now.

Your host

I’m Aymane Hitmi. I build with AI the way an agency builds with people.

I was the CTO of a US marketing agency, and I led a team of 12 developers. Today I build solo, with AI as my team. The habits that kept agency projects on track are exactly what this week teaches: a brief before a build, context before prompts, a test before trust, and one owner for every system.

I’ll be on every live hour, every hot seat and the clinic myself.

Former CTO of a US marketing agency.
Right now I’m building Misclo, my own fitness app, in development, launching 1 Jan 2027.

More about me

45+digital projects managed
12developers in the team I led
2software products I built and shipped

The lab

First cohort. No testimonials yet. Here’s what I build with the same method instead.

I won’t invent participant results. After this cohort, Builders who agree will appear here. Until then, judge the method by what it produces in my own lab, and by the free tools you can open today.

Misclo desktop homepage on staging: two stylised 3D characters beside the headline and one action.

MiscloIn development

My own fitness app, built with AI

An immersive marketing site and a WebGL 3D body map with 228 muscles and bones in 21 muscle groups, built with AI (model by Alex Lashko). In development, launching 1 Jan 2027.

Theo, a stylised 3D fitness character in orange shorts, rendered in a soft grey studio.

Theo, a 3D trainer character

Measured, not eyeballed

A Tripo AI mesh, rigged and animated in Blender with AI-written tools, measured against references rather than eyeballed.

The content engine

One finished reel, end to end

A research, render, voice and compose pipeline that produced one finished 17.6-second reel.

17.6sone finished reel from the content engine

Open the free blueprint See the lab

Why is this free?

Because it’s the clearest way to show you how I work.

I sell build work, a $970 Personal AI Audit, and soon a membership, Labgenz OS. Some people who build with me for a week will want help afterwards. Most won’t, and that’s fine.

So here is the deal, in plain words. Days 1 to 4 are building, with no pitch. In the last part of Day 5, after the build, I explain what I offer and who it fits, including when doing it yourself is the better choice. You never need to buy anything to finish the week or to get your deliverables.

Aymane

After Day 5

Three paths after Day 5. Doing it yourself is one of them.

Every path starts from the five things you build during the week, so the first step is the same for all three: save a seat.

01Both tracks
Keep building free
The free library, your Day Packs and, as a Builder, the Day 6 clinic. Many people should stop here, and that’s a good outcome.
Open the free library
02LAUNCH
Get the tool
Join the founding list for Labgenz OS, the project builder I’m making from this exact method. $46/month founding, locked while you stay subscribed, nothing charged until the Builder is live.
03LAUNCH · OPERATE
Get it built or audited
On LAUNCH, the Solo Launch Build, starting at $990, builds your launch assets for you. On OPERATE, a free 15-minute call tells you whether the next step is the Personal AI Audit & Implementation Roadmap ($970), a direct build proposal, or doing it yourself.

No countdown pressure. If a path isn’t right for you, I’ll say so on the call.

FAQ

Before you save a seat

When is it, and what time is it for me?

I publish the dates here once registration is confirmed. The early-access list hears first.

How much time does it take?

One live hour a day for five days, plus about 40 minutes on the homework path. The live build gets you about halfway, and each Day Pack marks what’s must-have and what’s stretch, so the 40 minutes gets you a submitted v1, not everything in the pack. Day 0 takes about 40 minutes before Day 1. On OPERATE, three days also start with prep you do before the live hour: about 30 minutes on Day 1, 20 on Day 3 and 10 on Day 4. Each pack says so at the top.

Is it really free? What do the tools cost?

The challenge is free. Day 0 runs entirely on free tiers. Each Day Pack lists its tools and costs before you start: where a free tier works, where a paid AI plan helps with long sessions, and which automation tools have a free tier or a trial. Make has a free tier with an operations limit; n8n’s cloud plan is paid after its trial, or you can host it yourself.

OPERATE Day 4 uses a pay-per-use AI API (Anthropic or OpenAI) inside the automation. You set a monthly spending limit before you connect it, and test runs usually cost very little; check the provider’s current prices. The key goes only into the automation platform’s credential manager. Upgrade anything else only when a free tier actually stops you.

Do I need to code?

No. LAUNCH builds its page with a no-code option or a single file you drag and drop. OPERATE builds its automation in Make or n8n from a step-by-step recipe, and the Day 4 pack shows where your API key goes. If you can follow a recipe, you can do this.

What if I can’t attend live?

Every Day Pack works without the live hour, and the homework path doesn’t depend on watching. Replay details are on the joining page.

What if the live room is full?

The live hour runs on Google Meet, and Meet rooms have a participant limit. If you can’t get in, that day’s Day Pack has everything you need, and the joining page says how to catch the session.

When is the Day 6 clinic, and who can join?

It’s for Builders: everyone who submits all five deliverables. It’s live with me, and we work through what your Day 5 review found. The date, time and length go out with the cohort dates and appear on the joining page.

Which track should I pick?

OPERATE if you’ve had paying customers in the last 90 days and you repeat work you could map. LAUNCH if you don’t have customers yet, or you want a new offer separate from your current business. You see both built live every day either way.

Payments or tools aren’t available in my country. Can I still do it?

Yes. Your Day 0 account plan flags anything blocked where you live, with a fallback, before Day 4 needs it. Payment links on Day 4 run in test mode only.

Will I make money from this?

I can’t promise that and I won’t. You’ll build real assets. Whether they earn depends on your market, your offer and the conversations you start.

What about my data and my customers’ data?

Never paste passwords, API keys, bank details or customers’ personal data into an AI chat. Nothing in the packs needs it. An API key goes only into your automation platform’s credential manager. Prospect research uses public business information only, no bulk scraping, and nobody’s personal details go into an AI chat. Every worked example is fictional and labelled. Automations run in test mode and draft for your approval until you decide otherwise.

Will my work be shown on screen?

Only if you say yes. Each submission asks whether it can appear on the wins wall, and hot seats are volunteers.

Is it really you on the calls?

Yes. If you see an AI version of me on Instagram, it’s labelled as AI. Every live hour, hot seat and clinic is me.

Will you sell me something?

In the last part of Day 5, after the build, I explain the audit, the build options and Labgenz OS, and who each one fits. Days 1 to 4 have no pitch. Nothing is needed to finish the week.

What happens after I sign up?

You land on the joining page: calendar file, the Day 0 setup and the Google Meet link once it’s published. Bring a laptop for the live builds.

Pick your track

Five days from now you could be looking at five things you built.

Starting from zero? Build the base of an AI-powered business. Already running one? Build the AI layer that runs alongside it. Either way, I’ll take the hardest step on screen first.

Free · Live on Google Meet · Two tracks

Save your seat

Registration isn’t open right now. The first tool is.

Open the free blueprint now and try again later.

Open the free blueprint

Five days. Five working pieces of an AI-powered business. Built live, alongside me.

You don’t leave with notes. You leave with files you’ll open again next week.

Day 0 to Day 6. One rhythm, every day.

Build Desk

Offer Card

Ops Map + AI Opportunity Scorecard

Business Brain v1 + 4-role AI team

Company Context Pack + first agent spec

Offer Page Kit v1: live page + lead magnet

Content Engine v1

First-Sale System v1 + prospect list

Automation Pair: speed-to-lead in test mode + back-office spec

30-Day Launch Plan + Solo OS v1

AI Operating Layer v1: dashboard, policy, 90-day roadmap

For Builders: we fix what your Day 5 review found

Same 60 minutes every day. Ten minutes of principle, the rest is real work.

Wins wall 5 min

Principle 10 min

Live build 25 min

Hot seat 12 min

Homework 8 min

Every day ships as a Day Pack you can use without me.

I’m Aymane Hitmi. I build with AI the way an agency builds with people.

First cohort. No testimonials yet. Here’s what I build with the same method instead.

My own fitness app, built with AI

Measured, not eyeballed

One finished reel, end to end

Because it’s the clearest way to show you how I work.

Three paths after Day 5. Doing it yourself is one of them.

Keep building free

Get the tool

Get it built or audited

Before you save a seat

Five days from now you could be looking at five things you built.

Registration isn’t open right now. The first tool is.