From an LLM chat to AI solutions

Who it is for

You lead a team or a department. Your people run into tasks every day where AI could help, but it is not yet clear how. Some have tried the chat and were not impressed. Others use it a lot and want to be able to do more. You want the whole team to move, not two enthusiasts.

Nobody on the team needs to know how to program. Without code it is even more interesting: people look at their daily work from the outside and see what can safely be handed to AI.

Works for any team:

• Accounting
• Manufacturing
• Finance
• Sales
• Support
• Documentation

How the day goes

A rough plan. We start around 10 or 11 in the morning.

1. 1–2 h
   Start: the philosophy of working with AI

   Not which buttons to press, but how this work is built: what changes from tool to tool and what stays the same. These regularities will still hold for the models and agents that come next. We look at who uses AI how today, go through the terminology and the basic toolkit.

2. Choosing a task

   I help each participant pick an idea: an automation, a service or a small app for their own work.

3. 45–60 min
   Break
4. 30–40 min
   Demo

   I build an app or an automation in front of you, from the task to the result.

5. 2–4 h
   Working on the projects

   A looser pace, with breaks. I am in the room, helping each participant.

6. 1 h
   Final demo

   Everyone shows what they built.

Along the way I show the tools I use myself: Claude Code, MCP, agents. No lecturing, only what will actually be useful.

• Claude Code
• MCP
• Agents

Where it happens

It is a weekend, so one option is your office: I come to you. But another one works better: leave the city for a day, somewhere with a view of nature and everything needed for work — a table, a screen, internet. A change of scenery helps people look at their usual work from the outside, and that is the whole point of the day.

I can suggest a few places.

Your host

Sasha Bondar

I have not written code by hand for a long time; AI writes it. What matters now is different: notice the task, decide whether it is worth solving at all, break it into parts, design the solution, build it, figure out what we learned along the way, and redo it with that knowledge. Then fit it into the daily workflow. Notice there is no «write code» step in that list?

I am writing a book about this, Requisite Variety, I run workshops, and I see every day how AI changes what one person can get done. At the workshop I show this process on a real example.

What participants say

Participant reviews will appear here.

Format and price