The field is stuck between two worlds that never quite fit: modeling structure and modeling movement. AI raised the stakes — and made the split impossible to ignore. Breakthrough Modeling unites structure, movement, meaning and AI-readability in one model.
impasse → provocation → breakthrough → momentum → now what
For years, one thing was never made clear: the difference between data modeling (structure) and data movement modeling (transformation). It’s the hard part — and the classic modeling tools never solved it. Each tool claimed a slice; nobody held the whole.
Structure
Data modeling
Entities, attributes, keys, relationships. A world of its own.
Movement
Transformation modeling
How data actually moves and changes. A separate world — and where the tools fell short.
Meaning
Knowledge
Semantics, business meaning, AI-readability — bolted on afterwards, never integrated.
Everyone models pieces. Nobody models the whole. That’s the impasse.
Three separate worlds that never fit together.
Data modeling came back — because AI needs it
Data modeling felt like something from a previous generation. Then AI arrived, and one thing became clear fast: for AI, a solid data model is essential. Structure is back on the table.
But a data model is no longer enough. We have to model our knowledge. That’s knowledge modeling — structure, movement and meaning in one.
The breakthrough
One way of modeling that unites structure, movement and meaning in a single, AI-readable model — the breakthrough that pulls a fragmented field out of its impasse.
AI isn’t added afterward here — it’s built in. The model is readable by AI, and AI helps you model: generation and validation, both.
One model, richly connected — with AI reading and building from the inside.
Layer
The old world — fragmented
BTM — unified
Structure
Data model, in its own tool
Part of one model
Movement
Transformation / ETL, separate
Part of one model
Meaning
Semantics as a loose layer
Integrated — knowledge engineering
AI
Added on afterwards
AI-native: readable + models with you
Knowledge
Locked inside a tool
Open, config-driven, file = truth
And the process itself is stuck
It’s not only the model. Scrum processes cost far too much time: too many meetings, without clear progress or results. The way we work around the model is as fragmented as the model itself.
We need structure. Prepared meetings. A new way of modeling — one where even the knowledge-gathering process is modeled and structured, not just the data.
The example app shows it: the whole path from question to model to knowledge is itself a model. Structure the work, and the meetings get shorter and the results get real.
The model behind it
The metadata model underneath it is real and runnable: 268 tables across 16 subject areas — structure, transformation logic, temporality, quality, lineage, governance and knowledge, all as queryable data rather than code buried in tools.
Breakthrough Modeling, introduced at a major Dutch bank
The first place BTM met the real world was a sparring session with the data team of one of the Netherlands’ largest banks. Not a pitch — a working session: put the modeling impasse on the table, question every assumption, and model anew.
The setting fit the idea. The bank’s headquarters is a curved, structural building — structure you can see. Inside, the conversation was about the structure you can’t: where data modeling ends and movement modeling begins, why AI needs a solid model, and what it takes to model knowledge, not just data.
Structure you can see. The session was about the structure you can’t.
What came out of it: a real conversation — and a first request to turn the approach into something a team can run, not just a one-off session. The impasse is real, and people are ready to break it.
