By Paul Preiss, of Iasa Global
I have been thinking a lot about tools lately.
Mostly because every architect I talk to is trapped in one. You know the feeling. Your team bought a modeling tool five or eight years ago. Somebody configured it. The metamodel grew a little every year, a field here, a custom object there, until nobody remembers why “Application Component” has three subtypes and a status called “Pending Pending”. And now ten years of your work lives inside a data model only that vendor can read. Want to leave? Export to Excel and start again.
Worse… what do you do when you change companies and have to learn that the decision records, requirements, services and capabilities that you were building are now considered wrong? How does truth morph based on your executive committee?
So let’s talk about what changes when the tools market answers to the profession instead of the other way around.
Who decides what an architecture artifact is?
Right now the answer is the vendor or the company. Each tool decides what a capability is, what a decision record holds, what a context view must show. They call it a metamodel and they sell it as a feature. Then your company morphs it through politics and out pops a sort of twisted concept of the truth. Every one of them is a little different, on purpose, because the difference is what keeps you in the tool.
That is backwards. Doctors don’t let the company that makes the patient record system decide what a diagnosis is. Structural engineers don’t let the CAD vendor decide what a load calculation means. The profession decides what the work product is. The tool helps you produce it.
In the BTABoK the work products are public, and they are decided by practicing architects.
Reviewers rate the concepts against criteria fixed for that kind of content. Then a committer, a senior contributor appointed to sign off what the BTABoK publishes, makes the final call. The people doing that work are committers, chief architects, and board-certified architects. People who have actually used a decision record to stop a bad decision. People who have drawn a context view for a board that did not want to hear it.
That is a very, very different standard from “our product team thought this field would be useful”.
And when a canvas changes, it changes the same way. Anyone can suggest a change. It gets reviewed. If it is accepted, the person who suggested it gets the credit. Nothing enters the body of knowledge on one person’s word, mine included.
Why that makes a real standard for vendors
A standard is only worth something if two conditions hold. It has to be made by people the users trust, and it has to be open enough that nobody owns it.
The canvases meet the first one. The storage meets the second.
The BTABoK stores structural models in CALM, the Common Architecture Language Model maintained by FINOS, and stores behavioral models (state machines, activity diagrams, sequences) as CoDL concepts that can be published to CALM. CoDL also carries the canvases themselves. So each decision record, context model or bounded context map has one defined structure that any tool can read and write.
Think about what that does to vendor lock-in.
If your decision records are BTABoK decision records stored in an open format, the tool you write them in becomes a choice you make each year. You can move. Your team can use two tools at once. Your AI assistant can read the same records your tool does. The custom data model stops being a cage because there is no custom data model or at a minimum it is additive. And for vendors who are honest about it, this is good news. They stop spending money inventing a metamodel nobody asked for and start competing on what they are actually good at. Better editing. Better collaboration. Better analysis. Better integration with delivery.
BTABoK licensed and accredited tools
So here is what we intend.
Tools will be able to become BTABoK licensed, which means they can carry the canvases, viewpoints and concern cards as the BTABoK publishes them. And they will be able to become accredited, which means they have shown they read and write the BTABoK model faithfully, in the open formats, without adding private structure that traps the user’s work.
Accreditation has to mean something to the architect using the tool.
I want three things to be true when you see that mark:
The artifacts are the real ones. The canvas in the tool is the canvas the profession approved, at the version the profession approved, with no fields quietly renamed.
Your work leaves with you. Everything you author can be exported in CoDL and CALM and opened in another accredited tool, intact.
It follows the method when the method improves. When a canvas or a viewpoint changes through review, an accredited tool picks up the change instead of freezing you on whatever it shipped in 2026.
The details of the program are still being worked out with the community and with vendors. I would rather build it in the open with you than announce something finished and wrong.
Why the viewpoint library is the best place to design
We didn’t invent the viewpoint library. We collected it. The best views the profession has produced over decades were scattered across books, standards and frameworks, each with its own vocabulary. We brought them together in one place, gave them one structure, and put them through the same review as everything else in the BTABoK.
Most design tools give you a blank canvas and a palette of shapes. That is freedom of the worst kind. You spend the first hour deciding what to draw, and the second hour drawing something nobody asked for.
The viewpoint library flips that. Each viewpoint is a set of rules for describing one aspect of a system. It names the concerns that matter, who holds them, and what a good answer looks like. Each concern has its own card and its own canvas. Context, Functional and Information are published. The rest of the sixteen are coming through the same review.
So when you design in it, you are answering questions instead of drawing pictures. Where is the boundary? What crosses it? Who owns this entity? What states can it be in, and who designed them? And because every answer is stored as a record in an open model, the answers connect. A decision points at the concern that caused it. A view stays in step with the model behind it.
It is still hard. A viewpoint tells you what to ask. Answering well takes years, and no tool shortens that. But it is an authoring environment where the structure comes from the whole profession and the thinking stays with the architect.
Equal footing, for every architect and every tool
This is where a lot of standards go wrong. They get written for one kind of architect, usually the one closest to the people writing it, and everyone else squeezes in.
The BTABoK supports five specializations: business, information, infrastructure, software and solution architecture. Each of them deserves canvases and viewpoints built for their work, and each of them deserves tools that support them as first-class users. A business architect should not have to use a software modeling tool with a capability plugin bolted on the side. An infrastructure architect should not have to fake a deployment view in a whiteboard app.
And the same goes for the tools. Accreditation has to be the same test for every tool, ours included. If a commercial tool does a better job for infrastructure architects than Ergane does, I want architects to know that and use it. The standard belongs to the profession. No single tool gets to own it, and that includes the one we built.
So why did we build Ergane at all?
Fair question. If the point is an open standard, why is IASA writing software?
Four reasons.
Experimentation. You can’t design a standard for tools without building one. Every canvas, every concern card, every CoDL concept got tested in Ergane before we asked anyone else to support it. When something didn’t work (and plenty of things didn’t), we discovered the problems in our own code rather than in a vendor’s software. Ergane is where the method gets its hands dirty.
Formal support levels. Ergane is the place we prove what each level of support actually looks like. Reading the canvases. Writing them back. Storing models in CoDL and CALM. Letting an AI assistant propose a change that an architect approves. Each of those becomes a level a tool can be accredited at, and Ergane is the working reference for every one of them.
A basic tool for member teams. A lot of architecture teams have nothing. A wiki, a drawing tool, a shared drive full of slide decks. Ergane gives a Premium member, or a team on an organization plan, a real place to do the work on day one. It is meant to be solid and useful. It is not meant to be the only tool you ever use, and it never will be.
A learning aid. This one matters to me most. You learn a canvas by filling it in on a real system. Ergane puts the canvases, the viewpoints and the self-paced learning in the same place, so an architect working toward CITA practices the method on actual work instead of on a worksheet.
And why we are so excited about the other tools
Because one tool, ours or anyone else’s, is never going to serve every architect in every company. The profession needs a healthy market of tools that all speak the same language. That is the whole point.
So the licensing is set up to open the door, not to tax it:
Free for open source tools. If you are building an open source modeling or design tool, you can carry the BTABoK canvases and viewpoints at no cost.
Free for nonprofits. Universities, public bodies, community projects. Same deal.
Free for starting tool providers. If you are a small company building the next good architecture tool, you shouldn’t have to pay us before you have customers. Start free, grow, and contribute back.
Established commercial vendors pay a license, and that money goes back into the body of knowledge and the review that keeps it honest.
Every tool that joins makes the standard stronger. More tools reading the same canvases means your work moves further. More tools writing the same model means your AI assistant can read it no matter where you wrote it. And more vendors at the table means architects, finally, are the ones the market has to listen to.
I genuinely can’t wait to see what people build.
Why this matters to you
Today architects are an afterthought in the tools market. Vendors sell to procurement, to the CIO, to the platform team. We get handed whatever was bought and told to make it work. Our artifacts get shaped by sales roadmaps. Our language gets shaped by marketing. When the work product is defined by the profession and the tools are accredited against it, that flips. The market has to come to us. A vendor who wants architects to trust their tool has to support the canvases architects approved, in the formats architects can take with them. That is real influence, and it is collective. No single architect could ever get that from a vendor on their own.
And it reaches further than tools. AI agents are going to read architecture. Platforms are going to distribute it. Regulators are going to ask for it. If our models live in a hundred private formats, every one of those conversations happens without us. If they live in one open model the profession governs, we are in the room. What would change in your team if your architecture could move between tools freely? What would you write down if you knew it would still be readable in ten years? Which canvas do you think the profession has wrong today?
Tell us. Better yet, suggest the change and put it through review. That is how this works.
The viewpoints so far:
Context viewpoint: https://www.iasaglobal.org/browse/btabok/3.2/core-site/core/article/context-viewpoint
Functional viewpoint: https://www.iasaglobal.org/browse/btabok/3.2/core-site/core/article/functional-viewpoint
Views and Viewpoints: https://www.iasaglobal.org/browse/btabok/3.2/core-site/core/concept/views-and-viewpoints
