Your team already uses AIeach in their own way.
I bring AI into a company starting from a single process: we decide where it works, who checks the result and how the whole team uses it. Anyone can ask Claude for a text. The work is noticing when the answer is wrong and turning that into a method that holds even when whoever found it is on holiday.
Three ways of working with AI
From your process to the classroom.
Workflows and automations
There's a step your team redoes every week exactly the same way: reports, variants, format adaptations. I lighten it with the simplest tool that does the job. If the result can be measured, we agree how before starting.
AI coaching
Your team already uses Claude or ChatGPT, but results vary from person to person. Three-hour sessions on real work, until prompts and checks work for everyone. Automations are a separate project.
AI training
You need to train staff for the AI Act, or you want the team to stop working by trial and error. In class we compare several routes to the same goal and discard the ones more complicated than necessary. Between sessions the team tries it on everyday work, then we discuss it in class.
Where AI enters a workflow
A slow process doesn't get fast by adding a tool.
If nobody knows who decides and what gets checked, AI just produces more material to fix.
First I design the process, then I choose where AI is needed.
How I work on a pilot project
Four steps.
Idea
I look at how you work today.
Direction
I choose where AI helps.
Production
I test it on your material.
Finishing
I check the result and hand it over to the team.
What I work with
I choose the tool after understanding the step.
- Claude
- NotebookLM
- Perplexity
- Mistral
- DeepSeek
- Qwen
- LM Studio
- ChatGPT
- Gemini
- Higgsfield
- GitHub
Frequently asked questions
Why isn't it enough to ask Claude or ChatGPT?
For a single text it is. The problem comes when AI enters everyone's work: ten people ask in ten ways, nobody checks results the same way and someone pastes client data where they shouldn't. With you I decide where to use it, who verifies what and which data stays out, and I put it in writing.
How do you introduce AI in a company?
You start from a single process. I map how it works today and where the same steps repeat: that's where AI comes in. We test on real material, measuring where we can, and move to the next process only if the first one works.
Is AI training mandatory?
For companies and organizations that use AI systems, yes. Since 2 February 2025, Article 4 of the AI Act (EU Regulation 2024/1689) requires ensuring staff have a sufficient level of AI literacy for the AI they use. It sets no number of hours: what counts is that people know what the tool does, where it goes wrong and who checks the result. I run this training for companies and agencies, particularly in Emilia-Romagna and Lombardy.
Do you also work with agencies?
Yes. In the projects you pass me, AI enters production with the rules set beforehand. The work goes out under the agency's name.
Do I need a technical team?
No. I build AI workflows for the people who do the work every day, with no code to touch.
Does AI replace the team?
No. AI removes the repeating steps; decisions and checks stay with people.
What do you have at the end of the pilot project?
The process in writing, ready to use without me: a map of who does what and where AI comes in; for each AI step, the tool, what to give it and what should come out; a list of what to check before a result goes out, and who checks it. When the result can be measured, I also leave you the before-and-after comparison: how much time, how many steps. I hand it over after testing it on your material.
What's left out?
I don't develop software products, I don't set up tools in your accounts and I don't reorganize the whole company.
Tell me what you need.I'll tell you how I'd do it and what I wouldn't do.
A few lines are enough: what, by when, for whom.
In my reply you'll find where I'd start and how long it takes.