FOUNDER
I build systems in plain sight
I am the founder of Vuntus. I keep every system I build, every decision I make and every mistake I make visible — trust is built with traceability, not with claims.
I studied computer science in Warsaw, at PJATK. I tested most of what I learned at school in systems I was building at the same time; that is where the gap between theory and production closed.
I built my first serious systems for my own companies. What was at risk was not a client's money but my own money and time — so every decision went through “does it work?” rather than “is it impressive?”.
At Vuntus I carry the same discipline outward: I build every system while documenting the process on camera. This is not a showcase — it is the construction itself.
My reason is plain: most “AI” offers sold to small and mid-sized companies sell a tool, not a result. What interests me is not the tool but the system your team actually uses and that produces a measurable result.
Vuntus is a new company. I build every system while documenting the process on camera.
Systems I have built
Two systems I built before Vuntus are still running under the same discipline.
agentic-base
Infrastructure that lets several AI agents work together in the same codebase, with human approval and leaving a trail.
GrowthOS
An operating system that gathers sales and operations data in one place and makes growth decisions measurable.
How I work
- 01
I define the result before starting
I do not start work until what we will improve, and by what measure, is settled in writing. At the end of work with no measure, everyone understands something different.
- 02
I put what is out of scope in writing too
What we will not do is written as clearly as what we will. Surprises should end in the conversation, not in the contract.
- 03
I do not let critical actions run without human approval
The system suggests a decision and prepares a draft; the price, the send and the critical step are always approved by a person.
- 04
I do not promise what I cannot measure
If I cannot show a result with a number, I do not promise it either.
- 05
I go deep in one process
Not everywhere at once. Really fixing one process is worth more than touching ten of them on the surface.
- 06
If your team cannot use it, the system is not finished
The work does not end when the build ends. The project is not closed until your team has taken the system into its daily flow.