Generating ArchiMate with AI
ArchiMate is the notation where generation works best and needs checking most. Its relationship table is machine-checkable, so a generator can be held to the spec - and its derived relationships let a model produce something legal, plausible and quietly wrong.
8 min readArchiMate 45 of 5
01Why ArchiMate generates better than free-form
ArchiMate publishes a relationship table: for every pair of element types, the set of relationships that is permitted between them. It is exhaustive, it is machine-readable, and it is the reason generation is on firmer ground here than anywhere else in modelling. A generator constrained by that table cannot draw a business process realizing a node or an actor accessing a technology service - not because it was well prompted, but because those edges do not exist.
The layers do the second half of the work. Every element belongs to exactly one of business, application, technology and the rest, so "which layer is this in" is a question with an answer, and a model that puts a database in the business layer is visibly wrong to a reader who has never seen your organisation. Free-form boxes and arrows offer no such handle; ArchiMate is a language, and a language can be checked.
02The derived-relationship trap
ArchiMate defines derivation: a chain of relationships implies a weaker one between its ends. A node serves an application, the application serves a process, so the node serves the process - and the spec allows you to draw that shortcut directly. It is a genuinely useful feature when a view is deliberately coarse, and it is the single most common thing a generator hands back without saying so.
The damage is not notational, it is that the middle disappears. In the figure above, the application doing the work is gone: nobody can see what would break if it were replaced, and nobody can see that it exists at all. A model built out of derived edges reads as an architecture with no applications in it.
The fix is to ask for the hops. "Show every relationship directly, do not derive anything" costs one sentence in the prompt and turns a tidy picture into a model you can reason about. The full mechanism is in ArchiMate relationships.
Two smaller failures round out the list:
- Layer confusion. Business services and application services get swapped, because the names differ by one word and the distinction is about who the consumer is rather than about what the thing does.
- Volume.Ask for "the architecture" and you get sixty elements, which is not a model but a list drawn with rectangles. Ask for a viewpoint.
03What to put in the prompt
Four things. The first two are the same as for any notation; the last two are what make an ArchiMate prompt different.
- The viewpoint."A layered view" to start with - it forces every element into a layer and forces the layers to connect, which is where the arguments worth having are.
- Your own names.The processes and applications as your organisation calls them. A model that invents "Customer Management System" where you say "Siebel" produces a picture nobody recognises.
- No derivation. One sentence: every relationship direct, nothing collapsed. This is the highest-value line in the prompt.
- A stopping rule."At most twelve elements", or one business process and everything under it. Without a bound you get the list.
04Checking the result
Reach for it when
- Any edge that crosses two layers - it is derived, and the middle is missing
- Every element's layer: is it there because of what it is, or what it is named?
- Serving arrows: they point at the consumer, and reversing one inverts the architecture
- Element count: if a reader cannot hold the view in their head, split it by viewpoint
Reach for something else when
- Re-prompting for a single wrong edge - fix it and keep the rest
- Debating element colour, which carries no meaning the spec defines
- Asking for every layer at once on the first pass
- Keeping a view that answers no question, however tidy it looks
The first line is the one that matters. Every other item on the list is a five-second judgement; a derived relationship standing in for two real ones is the error that survives review, gets published, and is still on a slide two years later with the application it hides still unnamed.
05How this works in Archyno
The AI is given the same metamodel the editor enforces, which for ArchiMate means the relationship table itself. An illegal edge is not something the generator is asked politely to avoid - it is not in the set of edges it can emit, and the editor will not accept one afterwards either. What comes back is a model in the repository, not a picture of one: rename an application and the name changes in every view it appears in.
From there it exports to the ArchiMate Exchange format that Archi and BiZZdesign read, to a Sparx .qea file, and to PNG, SVG or Mermaid for a document. That matters more for ArchiMate than for anything else on this site, because enterprise architecture work is judged by people who use a different tool from the one it was made in.
In one line each
- 01ArchiMate's relationship table is machine-checkable, so generation cannot produce an illegal model.
- 02Legal is not correct: the table says a relationship may exist, not that yours does.
- 03The common failure is a derived relationship drawn as a direct one, hiding the middle layer.
- 04Prompt with a viewpoint, your own names, no derivation, and a stopping rule.
- 05Check cross-layer edges first - they are derived until proven otherwise.
- 06Generate into a model, or the result cannot leave the tool it was made in.
The UML side of the same question - where the rules are looser and the failures are different - is in generating UML with AI.
06Common questions
Can AI generate an ArchiMate model from a description?
Yes, and it is more reliable here than in looser notations, because ArchiMate publishes a table of which relationships are permitted between which element types. A generator constrained by that table cannot produce an illegal model, so the errors that remain are about your organisation rather than about the notation.
What do AI-generated ArchiMate models get wrong?
Mostly they draw derived relationships as though they were direct ones - a technology node serving a business actor, which is legal and hides the application that actually does the work. After that: elements placed in the wrong layer, and far too many of them, because a model asked for completeness produces a picture nobody can review.
Which ArchiMate viewpoint should I ask for first?
The layered viewpoint, because it forces the generator to place every element in a layer and to connect the layers, which is where the useful arguments are. Ask for a more specific viewpoint once you know which question the model is being built to answer, rather than before.
Can a generated ArchiMate model be opened in another tool?
Only if it was generated as a model rather than as a picture. A model exports to the ArchiMate Exchange format, which Archi, BiZZdesign and Sparx all read, so the work survives the tool it was made in. An image of a diagram exports to nothing and has to be redrawn by hand.
Related reading