A working automation
Built in the room on your real process. From Monday it works alongside you, instead of sitting in the slides.
In 16 hours your team picks a real process and builds an automation they use right away, no IT team required.
AI Adoption Sprint
You leave withAn automation built on your process, ready Monday morning
Yempik’s partners
Your people already use ChatGPT for emails, documents, and reports, often with no guidelines. The problem isn’t the tools: it’s knowing what AI can really do, what it can’t, and where it saves time instead of wasting it.
That’s why we start from the process, not the tool. And when AI isn’t the answer for a process, we say so: saying it, when it’s true, saves time and budget.
Starting in 2025, the AI Act (art. 4) requires companies to ensure an adequate level of AI literacy for anyone using these tools at work.The course builds that foundation. Legal or compliance advice stays with the dedicated professionals.
Current state
75% of knowledge workers already use AI at work; 78% bring their own tools, often with no company guidelines.Fonte: Microsoft and LinkedIn, Work Trend Index 2024
Built in the room on your real process. From Monday it works alongside you, instead of sitting in the slides.
What to automate next, in priority order: impact, effort, and where to start.
The team understands risks, limits, and rules of use: what to share with AI, what not to, and how to verify it.
Every module produces a piece of the result. The theory is there, but it serves the build: the throughline is your process, from start to finish.
What AI can really do, what it can’t, and the minimum rules to follow. You pick the process you’ll work on for the whole course.
Your process, chosenYou standardize the process and build the first automation in the room, step by step, without writing code.
Your automation, first versionAI works on your real documents, safely: what to share, what to keep out, how to check the results.
A reliable automationYou measure the time saved and plan what to automate next, in priority order.
ROI and a 90-day roadmapNo technical background needed. If you can use email and documents, you’re ready.
No code, no servers, no IT department. The tool runs on the person’s computer: the automation you build in the room is the same one you use every day, on emails, documents, sheets, and reports.
The method, though, comes before the tool: it starts from the process and stays valid with any AI assistant, today and tomorrow.
Skills, not just knowledge
From slides to a second brain
Pick a result: the method is always the same.
Complete decks, from structure to content.
One single method, with Claude Cowork or any AI assistant, today and tomorrow.

Founder, Yempik
For over 6 years he has built and shipped AI agents and automations to production for companies across different sectors: voice agents for call centers, automations on email and documents, integrations with business systems.





AI Engineer, Yempik
He builds agents, knowledge bases, and workflows that go to production, designed to stay traceable and compliant. He has led generative AI projects inside companies since 2023.



Everything we teach in the room is published here, for free: eight practical guides, from the first process to automate all the way to ROI. Read them before you call us, if you like.
Custom quote
The figure depends on the number of people and the location: we tell you on the first call, before any commitment.
Next dates coming soon
Individual seats, several companies in the same edition. Leave us your email: we’ll let you know when the first date opens.
The flagship format is the 16-hour Sprint, over 4 half-days. On request: shorter versions for leadership or extended ones for several teams.
No. The course is designed for people who manage processes and people, not for technical staff. If you can use email, documents, and spreadsheets, you have everything you need.
The ideal format is between 6 and 12 people: enough to work on several processes, few enough to follow each one all the way to an automation that works. For larger groups we run multiple editions.
Whichever you prefer. In-house we’re glad to work on site, but the course works just as well online and live: you still build, each person on their own computer.
16 hours, split into 4 half-days. That way the team’s work never stops for a full day, and between one module and the next there’s time to try things out on real work.
A working automation built on your process, the roadmap for the first 90 days, the map of use cases for your functions, a checklist of risks, and the foundations for an internal AI policy.
It’s designed for exactly that. You work with no-code AI tools that run on each person’s computer: no servers, no installations, no technical department.
No IT project, no complicated installation. In the room you use Claude with Cowork, the no-code tool that runs on each person’s computer: you need a paid Claude plan (from about €20 per person per month, Cowork included; that’s the cost of the tool, separate from the course). For in-house courses we recommend Claude Team, the business plan where data doesn’t feed training: you activate the accounts, they’re agreed in advance, and they can be canceled at the end of the course.
Yes, and it’s part of the course, not a side note. In open-enrollment editions you work on sample data, never on your real documents. In-house, when real data is used, you use a Claude business plan where inputs aren’t used to train the models, agreed with your IT team. And the "Data, documents, and control" module teaches exactly what to share with AI and what to keep out.
It depends on the number of people and the format. We tell you the exact figure on the first call, free and with no commitment, before any signature. No hidden costs.
Fill out the form: we reply within 24 hours with a proposed intro call. Free, 30 minutes, no commitment.