Your start
No prior skills needed.
Start with a task you know. Excel, for example.
Into the world of data.
01 / Something bigger begins
What are you missing
by not starting?
Training · AI deployments · Websites
02 / It's already happening
See how much we've drilled: analysis, research, documentation and code produced with our AI agents.
How many of your ideas are stuck for lack of time?
How we count ↓one person's work
27× more than two people would do in that time
Estimated output equivalent.
Result depends on the conversion factor.
03 / More room for you
Record a voice note, the agent replies in text. Work wherever you are: at a desk, on the road, at the wheel.
Do what you want, where you want. The decision always stays with you.
Swipe the conversation or pick an agent.
A demo. Results need checking.
04 / Our mission
We want to grow digital skills in Poland and help people use their potential with AI.
We teach at different levels, at a fair price. From Excel to your own AI solutions.
Meet the people behind itYour start
Start with a task you know. Excel, for example.
Your practice
Practise, compare results and learn to ask sharper questions.
Your independence
From data work to agents and your own solutions.
Potential is in people.
Let's help it work.
05 / The people behind it
Dominika and Bartek. She teaches people to work with data, he builds tools with AI agents. Click a person to find out more.
ISO 9001:2015 · certificate no. 26EQRB82 · valid until 28 Sep 2029 Pick my route
Data and training
Teaches Excel, Power BI and data analysis. Shows how to use AI in everyday work.

Technology and AI agents
Builds apps with AI agents. Combines models, his own hardware and tools into working solutions.
06 / Your move now
Learn to do it yourself,
hand us the task or visit the lab.
Still finding your direction? Let's discuss your task ↗

Expert in Excel, Power BI and data analysis training. 17 years working with company data, over 3000 people trained.
At Numerika she makes sure AI knowledge is practical, testable and in plain language: reports, formulas, file work, plus AI that really shortens the working day.

Expert in building apps and working with AI tools, including vibe coding: building software in close collaboration with AI agents.
He runs Numerika's technology: an in-house compute cluster, a team of agents with company memory, and the apps and sites built on them. This site came from that lab too.
The counter shows an estimated equivalent of output volume. It is a comparison under stated assumptions, not a measurement of time saved.
Snapshot from 24.09.2026, 20:17 UTC: 231 218 lines of code and 1 888 485 words. Both variants assume 1760 working hours per year. The main screen shows the first variant.
27 times more than two people: from 1 February to 24 September 2026, 235 days passed, or 0,64 of a year. Two people working for that time make 1,29 years of work. 34,9 years divided by 1,29 gives 27. In the second variant: 14,8 / 1,29 = 11,5.
Word and line counts do not measure quality, time spent checking and correcting, or business impact. We do not convert them into guaranteed client productivity.
For a specific task, savings require comparing working time before and after, including checking the result.