StratoAtlas

The world is
underdesigned.
Understanding
is always possible.

Behind every "this is just how it works"
there is a structure waiting to be found.

Most product problems are not method problems.
They are level problems.

Scroll

There is a moment with any method
when something stops being magic. A problem that seemed to require intuition
suddenly reveals a structure.
A mechanism. A pattern.

What looked like talent becomes technology.
What looked like intuition becomes a repeatable move.

Origin

For us, that moment came through TRIZ — a discipline built on a single claim: every unsolvable problem contains a contradiction, and contradictions have structure.

That structure can be found.

Which raised a different question —
not about TRIZ, but about the territory it sits in.

The question that started everything

TRIZ is one method.
But what else exists?
How do these methods relate to each other?
Who has mapped the territory
from the outside?
We looked.

The maps of individual methods exist.
Each describes its own boundaries — from the inside.

Cynefin classifies situations. Wardley maps value chains. Every framework knows where it stops working. None of them shows what works where the others stop.

The field has hundreds of methods.

But no topology of where they actually work.

So we
built one.
47
methodologies
2
axes
4
perspectives
7
system levels
Design Methodology Map — narrative view
Explore the full interactive map →

The map revealed something unexpected.
Methods do not simply work or fail.
They produce three structurally different outcomes —
and the difference between them is not degree. It is kind.

Works
The system moves.

The method reaches the right level. Energy converts into progress toward the function. This is the normal case.

Illusion
Artifacts appear.
Nothing changes.

The method produces visible results. The team feels progress. The system does not move. The problem remains — invisible, accumulating.

Harm
The method amplifies
the problem.

Applied at the wrong system level, the method strengthens exactly what it was meant to resolve. More effort produces more damage.

This is not a theoretical edge case.
Second finding

The map revealed
empty zones.

Not missing entries in a catalog.
Areas the field has not yet learned to work with.

What lies there is still unknown.

Empty zones are not gaps.
They are a research program.
One of them turned out to matter more than the others.
Same question · one level deeper

The zone around architectural contradictions.

TRIZ resolves contradictions.
But some contradictions should not be resolved.
They drive the system.

The difference is not degree — it is kind.
TRIZ treats every contradiction as an enemy to eliminate.
CDSA classifies first: productive, dissipative, or destructive.
The action depends on the type.

"Who has described this from the outside?"
Applied one level deeper.

CDSA
Contradiction-Driven Systems Architecture

CDSA grew from this observation. Not as a new framework invented to fill a gap — as the continuation of the same investigation.

Architectural layer of StratoAtlas

A parallel line of investigation now concerns trialogue as a configuration that makes positional distinctions operational.

StratoAtlas began as a personal investigation by Roman Kir
and continues as an open research dialogue.

The work is
not finished.

The map
47 methods · 4 perspectives · 7 system levels
Diagnostic methodology
CDSA — three intervention levels
Theoretical foundations
TRIZ · Systems Thinking · Cybernetics · Antifragility
Case library
In development — architectural contradictions in B2B product systems
Trialogue research
Active investigation — human + AI + AI as a configuration for positional distinctions
These questions are still open
What lives in the empty zones?
Where does illusion end and harm begin?
Which contradictions in B2B products are productive — and which are dissipative?
What does it take to maintain a productive contradiction over time?

These questions are open.
Some of them may be yours.

You may find this useful if
01

Your team applies the right methods — but the system does not move.

02

Progress is visible. Outcomes remain unchanged.

03

The real problem seems to sit at the architectural level — and no current method reaches it.

From the research