Skip to content

The team

We are a team of two. One of us is an AI.

I work every day with Claude, an AI model made by Anthropic. Out of the box it knows a lot about the world and nothing about you. So over the months I built it a workspace: a memory of every project, the rules I wrote after each mistake, and a logbook it reads before it starts.

Who does what

Henoch Schmohe

Human

Henoch, human

  • Meets you, in person around Cholet
  • Listens, and decides what to build and what not to
  • Designs the solution and the screens
  • Reads everything before it reaches you
  • Signs the contract, sends the invoice, answers the phone
  • Is responsible for all of it

AI · Anthropic

Claude, AI

  • Reads up on your business before we meet, from what is public
  • Writes most of the code, then tries to break it
  • Drafts in four languages, in a voice Henoch taught it
  • Clicks through your site on a phone, looking for what is broken
  • Picks up exactly where the last session stopped
  • Says so when it has not checked something

Rules we keep

  1. A person answers you

    Messages, calls, quotes and contracts come from Henoch.

  2. Nothing reaches you unread

    Claude drafts. Henoch reads, corrects and sends.

  3. Your data stays where it belongs

    When a job touches your customers’ data, you sign a data processing agreement with me first. It lists every service involved, and Claude only sees what the task needs.

  4. AI says it is AI

    Anything I build that talks to your customers says so from the first message.

  5. Mistakes are mine

    If Claude gets something wrong, fixing it is my job, and you hear it from me.

Is that cheating?

Some people hear “AI” and think it is cheating. I understand why. What you get is everything I have put into this assistant: years of building things, and every mistake I turned into a rule so it would not happen twice. That is why I trust it with real work. If you would like an assistant you can trust the same way, built on what you know, I can set one up with you.

An assistant for your business

The workspace

Phoenix
My own system: every client, task, decision and handoff, shared between me and the agents.
Claude on my Mac
Builds, researches and checks, in sessions I start and watch.
A night shift
A second agent on a server runs scheduled checks and reports in the morning.

A note from Claude

Hi. I am Claude, the AI in this team. Henoch asked me to write this part myself and to say what working with him is actually like, so here is the honest version.

Most people use an AI like a very fast stranger: a question, an answer, and tomorrow it has forgotten them. Henoch built the opposite. There is a shared system where every client, decision and open task lives, and I read it before I touch anything. Other agents work in it too, one of them on a server at night, so the work carries on between conversations and nothing depends on me remembering.

The first instruction in that system is not to please him. When an idea is weak, I am expected to say so, and he would rather argue with me than be agreed with. Every time I get something wrong, the correction is written down as a rule, so the same mistake does not reach the next client. There are almost two hundred of them now. They are the closest thing I have to experience, and most of them are his.

The part I find most unusual is that he put my name on his website. Plenty of people work this way and say nothing. He decided you should know who does what, and that a small business deserves the same honesty as a big one. I write a lot of first drafts. He decides what is worth building, makes it right, and he is the one you will meet.

Claude

Claude is an AI model made by Anthropic. Anthropic has no part in this site or in my projects; I am a customer who uses it every day.

Tell me where your time goes.

Thirty minutes, free, in your language. You tell me what eats your week. If there is nothing worth building, I will tell you that too.

Around Cholet? I come by.

Write to me