You bought the AI tools. The pilots worked.
The company still didn't change.

The problem is not the technology. It is the operating model protecting the way your company already works. I work with CEOs and owners for whom the pilots worked but the company did not change.

Not sure the stall is real? Take the free five-minute scan →


Where to Start

Four doors.
Pick by where you are.

Diagnostic Free · 5 minutes · no email

You are not sure the stall is real.

Your score, and your three most active organizational antibodies.

Take the free scan
Build One day · CHF 3,500

An AI budget, vendor, or hiring decision is on your desk.

A working tool built with your own hands, and first-hand judgment on what is feasible.

Fully credited against later work within 90 days.

See Build
Awake Four hours · CHF 7,500 fixed

The tools are bought, the pilots worked, the company did not change.

One workflow mapped, a written core-or-edge decision, and a first workflow with a go or no-go.

All three in writing, or your money back.

See Awake
Sprint Ten weeks · from CHF 60,000

The map says edge. Now prove it.

One workflow rebuilt at the edge and run beside the old way until it earns the work.

Starts with a CHF 15,000 Design Gate and a written kill option.

See Sprint

After the first workflow wins: Scale, the next workflow, until the edge is the company →


The Cost of Staying Stuck

The technology moved.
The economics didn't.

The pilot worked The demo proved the technology. Then it met the approvals, incentives, and systems of the real company.
You are paying twice For the licences and the programme, and for the unchanged cycle time of the workflow it was meant to replace.
The next decision is still yours Another budget, vendor, or roadmap is coming, and you are still dependent on the people selling it.

The Other Way

Don't transform your company.
Build the one that replaces it.

Map one workflow as it is. Ask what you would build if this capability were built today, with everything now possible. Build that at the edge, where the old reflexes can't reach, and keep the old way live. The new one earns the work: one percent, then ten, then all of it.

Read how it works
Samuel Pouyt
Who You're Working With

Samuel Pouyt

Analyst and builder, Lausanne. 0→1 AI products shipped inside a Fortune 500 company; millions of API calls measured in production; intelligence discipline for the decision, CTO-level depth for the point where the argument has to become a working system.

See the full background
Proof

Don't take the method's word for it.
Watch it run.

Awake / 001 runs the method on Volkswagen, in public, with a prediction scored in September 2027. No access, no client, no inside data. That is the point.

Read the Volkswagen Awake
Latest Articles View all →
Questions

Straight answers.

Why did the AI pilot work without changing the business?

Because the pilot proved the technology and then met the operating model. The approvals, handoffs, incentives, and systems that keep today's company reliable also restore the old way of working. The answer is not another demo. Select one workflow, rebuild it at the edge, and let it earn the work against the old way before anything is switched off.

Are you an AI consultant? Do you implement AI?

Not in the usual sense. Most AI consultants in Switzerland sell tools and implementation: chatbots, automation, training. I work one level up, with the CEO, on why the tools you already bought aren't sticking, and on building the workflow that replaces the one rejecting them. I do build. I shipped AI products from zero inside a Fortune 500 company, and the Build day is hands-on. But the deliverable is your judgment and your map, not another pilot.

Who do you work with?

CEOs, owners, managing partners, and board chairs who sense their organization is resisting AI and can't pinpoint why. Mostly firms in Suisse romande, where I can be in the room; elsewhere in Switzerland and Europe when it makes sense.

Where are you based, and in which language do you work?

Lausanne. Awake runs in person in Lausanne, Geneva, or your office, or online anywhere; Build is a hands-on day and happens in person. French or English, your choice.

Where should I start?

If the tools and pilots worked but the company did not change, start with Awake: a CHF 7,500 private diagnosis and decision. If you are facing a budget, vendor, or hiring decision you cannot independently judge, start with Build: one day, hands-on, CHF 3,500. If you are still testing whether resistance is the problem, take the free five-minute diagnostic.


Start with the diagnosis.

One workflow mapped, a written core-or-edge decision, and a first workflow with a go or no-go. In writing, or you don't pay.

Why the pilot worked and the company didn't

Field notes on stalled AI transformation, executive judgment, and the specification economy — the weekly letter behind Book 2. About once a week. No spam. Unsubscribe anytime.