The Praxis Programme
Eight sessions.
Your team leaves AI-fluent.
Your staff already use AI, and it saves them nothing: ask, fix, ask again. But your team knows its own work inside out, and that's where we start. First we set up your AI around your business: the rules, what good looks like for each job and what to hand over. Then we build on that setup together, on your own work, through real conversations and worked examples, up to the tools, skills and automations that do the work without you.
After eight sessions your team is AI-fluent: they can find their way around all of it and build anything they can describe. No coding background needed. You leave with the charter, the scorecards, the briefs and the checks, and the jobs that used to repeat, handed over.
Bring the one job that's bugging you. In an hour you'll get started with our scorecard and policy tools, and leave with your first artefact. No slides, no pitch you didn't ask for.
You and up to four of your team · Online or in person
Most teams start with the tools.
That's why nothing sticks.
Your staff have typed questions into ChatGPT or Claude. Some answers were useful. Most needed fixing, and the time saved went on the fixing. The problem isn't the tool. Nobody set the rules, defined what good looks like, or decided what to hand over. This programme does those three first, on your own work, then builds on them.
Who This Is For
Your team already uses AI
And it takes as long as before, because every answer gets checked, fixed and asked again.
A job that repeats
The quote, the report, the weekly export. Work a machine could do, if someone set it up properly.
You can't see what goes in
Client files and personal data, pasted into tools nobody vetted. You want a rule before it becomes a problem.
Nobody needs to code
Nobody on your team has to become a developer. You want the work done once, not a new hobby.
What You Will Be Able to Do by the End
Set the rules: what goes in, which tools, and what never leaves the building.
Judge AI output against a KPI for each job, not a feeling.
Decide which jobs to hand over, and what each needs to run without you.
Build the tools, skills and automations that do those jobs, by describing them in plain language.
Keep building on your own after the programme ends.
The Eight Sessions
Weekly, over roughly two months. The first four set the foundations: the rules, the standard, the decisions, the briefs. The last four build on them.
Rules first
What goes in, which tools and connections are allowed, what never leaves the building, and what happens when the rule is broken. We set up your assistant safely while we write it.
You leave with: your AI Use Charter
Define good
Pick the jobs that repeat. For each one, write down what good output looks like and how you'd score it. These are your evals: the standard every piece of AI work gets judged against.
You leave with: a KPI Scorecard per role
Decide what to hand over
Which jobs AI should do, which it shouldn't, and what each one needs to run without you: the inputs, the rules, the check at the end.
You leave with: the list of what AI takes over
Brief it once
Turn each job on the list into standing instructions: the audience, the constraints, an example of good. Written once, read by the assistant every time, so nobody explains the same job twice.
You leave with: standing briefs for your top three jobs
Tools
Connect the assistant to your files and the apps you already use. Build the first small tools that do a job end to end, with you in control of what they can touch.
Check it before it ships
Score the output against your KPIs. Where it falls short, fix the brief, not the draft. Agree who signs off and what gets checked before anything leaves.
You leave with: your Verification Protocol
Skills and automation
Package what works into skills the whole team can call, and schedule the jobs that should run without anyone asking.
Agents, and a clean handover
Your first agent: a job that runs on its own, on your data, checked against your scorecard. Then hand it over cleanly, so it is still changeable in a year.
What You Finish With
The rules, the scorecards, and the jobs handed over.
Your charter, a KPI scorecard for each role, standing briefs and checks for your top three jobs, and the tools you built on them. A certificate records what you made, not just that you turned up.
A note on honesty, because it sets the right expectation: you will get most of the way there yourself. On anything complex, the last stretch you finish by hand or hand off. Even getting most of the way is a large saving on how the work is done today - and that is the saving we are after.
Why a Person, Not Another Video
Tailored to your actual work
Not a generic example you have to translate. We start with the task that eats your week and build from there.
We meet you where you are
No burying you in tools you have never heard of to look clever. Every session starts from where you actually are.
The fiddly parts are where a guide earns their keep
Installing the tool, the first setup, the moment something breaks: that is exactly the part the free videos skip, and exactly where most people quietly give up.
How It Works
Format
You and up to four of your team. For one to one, take The Solo Fast Track. Either way it is built on your own work, not a syllabus.
Your use cases
Before we start, send the jobs you keep repeating: the quote, the report, the weekly export. We work on those from the first session. The group evenings use generic examples. This doesn't.
Cadence
One session a week, 90 minutes each, over roughly two months. A short practice task between each session.
Where
Online, via Teams or Google Meet. In person on request.
What you need
A laptop - Mac or Windows. An account with the AI assistant of your choice. We set up everything else together in the first session.
On the choice of assistant
We teach with Claude as the worked example, because it is the one most people arrive with. It is not a requirement. The same method works with Mistral, with OpenAI, or with open-weight models running entirely on your own machine, which is the right answer when the work cannot leave the building. You choose the assistant; we set it up with you in the first session.
Everything You Get, and What It's Worth
Each piece priced as if you bought it on its own. Then the guarantee, then what you actually pay.
The AI-Fluent Team · what you get
AI Use Charter
What may go into the tools, which tools and connections are allowed, and what happens if the rule is broken. Written in session one.
€1,400
KPI Scorecard per role
What good output looks like for each job, written down, so feedback stops being "make it better".
€1,400
Hand-over list
Which jobs AI takes over, which it doesn't, and what each one needs to run without you. Decided in session three.
€1,000
Briefing Library
Standing instructions for your top three jobs, so nobody writes the same brief twice.
€2,300
Verification Protocol
How output is checked before it ships, and who signs it off.
€900
Eight live working sessions
On your own work, not exercises. Nothing is homework that could be done in the room.
€3,700
Bonus: The Field Guide
For "my team won't remember this": every principle on one page per person.
€200
Bonus: 30 days of async review
For "it won't stick once you're gone": one piece of work a week, reviewed.
€700
Bonus: 90-day check-in call
For "we'll drift back in three months": we look at what slipped and fix it.
€400
Total stated value
€12,000
The 30 days of async review is included for anyone who starts by 30 November. After that it is an extra.
The guarantee
Pick one recurring piece of work before we start. By the last session your team produces it with AI to the standard on your own scorecard, or I keep working with you at no charge until they do.
The only condition is that you attend the sessions and answer the questions.
What you pay
The AI-Fluent Team
€4,500€2,000
Back-to-work price until 30 November
Eight 90-minute sessions for you and up to four of your team. Everything in the stack above.
The Solo Fast Track
€2,000€1,000
Back-to-work price until 30 November
Four 90-minute sessions, private and one to one. The same artefacts, scoped to you, with the Field Guide and 30 days of async review. Credited in full if your team follows.
Refer a friend: €100 off your programme for every person you refer who enrols, up to €500.
Referral credits apply to courses taught by participating trainers.
Start with the free 60-minute session: proof on one of your own jobs before you spend anything. When you book the course, a 15-minute call to say hello and set the dates comes with it. Both programmes are invoiced, ex VAT. We also run free sessions from time to time. They are general rather than built around your use case: see Events.
Running the tools you build costs about €19/month for a subscription to a frontier AI system (Claude, Mistral or OpenAI), or nothing at all if you run open-weight models on your own hardware. The tools themselves run on your machine, with no per-use charges.
Common Questions
On what this will not do
It will not make you a software engineer in eight weeks.
It will make you someone who can build genuinely useful things, and who knows when a job is big enough to call in a developer, and how to brief them when you do. That is a more valuable place to stand than it sounds.
Book your first course
Tell me about your team. Then we pick the dates.
A few details so the first session starts on your work, not on introductions. Once they're in, you book a 15-minute call to say hello and set the dates.
Not ready to book? Book the 60-minute session, free · Or email directly: daniel@tutto.one
What comes next
Become a Praxis trainer.
Teach it, and earn from it.
Four sessions on top of the programme turn you into a trainer. Clients you bring in yourself: you keep 80% of the tuition they pay. Clients Tutto brings in: you keep 60%. I find and organise the clients with you, you deliver the sessions, and a hub is behind you for the hard jobs.
Trainer track
€1,000
Four sessions
You keep
80%
on the clients you bring in yourself