What Is Prolaborate? The Sparx EA Stakeholder Portal, Now With AI

If you have ever exported a clutch of diagrams into a slide deck the week before a steering committee, you already know the job Prolaborate exists for. The job is getting current architecture in front of an audience that has no intention of learning a modeling tool, and getting their reactions back. This piece walks through what an architect can now get done with it, how each task is done today and done AI-augmented, and which other tools help with neighboring jobs.

What an architect can now get done with Prolaborate

Put the architecture in front of people who never open the client

Today. The architect exports diagrams into slides before each cycle, the slides go stale within weeks, and every question about them comes back by email.

With Prolaborate. Selected diagrams, matrices, lists and dashboards are published to a portal that follows the model as it changes. Stakeholders search, follow links between related elements and leave comments without touching the model. Role-based access decides which audiences see which views.

The skill. Deciding what each audience needs to see and what you want them to do with it: a program manager checking what is in scope, a process owner signing off a model, a review board reading a proposal. Publishing everything is easy and helps nobody. For how this plays out for portfolios, see Application Portfolio Management.

Tools that help. Architects who work with Claude and AI Power Tools for Sparx EA can also have the AI build and read Prolaborate dashboards through the browser. The architect states what the audience should see, and the dashboard building is no longer the architect's job.

Make a long discussion or a dense diagram quick to take in

Today. A review thread grows to dozens of comments, and a new reader either reads all of it or asks the architect for the story. A large diagram gets the same treatment.

AI-augmented. With Prolaborate 6 with AI switched on, a reader can ask for a summary of the discussion or the diagram. Sparx also describes it as answering questions about the model.

The skill. Knowing what a good summary of this particular diagram would contain, asking for that, and checking the summary against the diagram before anyone acts on it. A summary reads as finished whether or not it is right, so the checking habit matters more than the button.

Get descriptions written, and written for the right reader

Today. Descriptions are written by hand, one element at a time, so most stay thin or blank, and the ones that exist are written in modeler's language.

AI-augmented. Prolaborate 6 with AI can generate a description and rephrase existing text. The architect says who the description is for and what it has to cover, then reviews the result.

The skill. Stating the audience and the purpose. A description of one system for an executive sponsor and one for a delivery team are different pieces of writing, and the AI writes what it is asked for.

Let stakeholders ask instead of email

Today. Every question about the architecture goes to an architect and waits for a free hour.

AI-augmented. A stakeholder in the portal asks a question about the model and gets an answer there. Someone on the architecture team decides to switch the AI on, connects the Kernaro AI Hub license and the organization's LLM, and decides who can use it.

The skill. The answer is only as good as the model behind it and the question put to it. Architects who watch what stakeholders ask learn where the model is thin and what to model next. The portal reflects the model for better or worse, so an AI answering from a patchy model gives patchy answers.

Which helper fits which job

Prolaborate is one of several tools around Sparx EA, and the right one depends on the job and the people doing it.

• Stakeholders already work in Prolaborate and need help reading what is there. Prolaborate 6 with AI.
• Stakeholders need to ask open questions across the whole repository and the other tools your team uses. Kernaro AI Hub, a web application that gives stakeholders a conversational view of the live model.
• An architect needs to build or update the model itself. Kernaro Assist inside EA, or Claude or GitHub Copilot with AI Power Tools.
• Stakeholders live in Teams, Word and Outlook. Microsoft 365 Copilot, read-only, from a Power BI model built on your EA data.

The companion piece Prolaborate or Kernaro AI Hub works through the first two choices, and AI tools for Sparx EA sets all six options side by side.

Who uses Prolaborate

The primary audience is business stakeholders who need to engage with architecture but do not work in Sparx EA.

• Program and project managers who need to see which applications are in scope for a transformation, what integrates with what, and what the target state looks like.
• Business process owners who review and sign off process models without learning a modeling tool.
• Governance committees and architecture review boards who want to read architecture views without sitting through a walkthrough.
• IT management who want a recurring view of the application portfolio without navigating the EA client.

They get access without licenses, training or a client installation, and architects keep control over what is published and when.

Licensing and procurement

Prolaborate is licensed directly from Sparx Systems. It is a separate product from Sparx EA and needs its own licensing, with pricing that varies by user count and deployment model (cloud-hosted or self-hosted). The AI features need a Kernaro AI Hub license on top, plus your own LLM, and you pay that provider for usage.

Sparx Services does not resell Prolaborate or Kernaro. If you are evaluating either, procurement runs through Sparx Systems or an authorized reseller. Sparx Services is independent of Sparx Systems Pty Ltd. What we do is teach architects to get the work done with these tools, which is separate from the commercial terms between your organization and Sparx Systems.

Can Prolaborate replace PowerPoint architecture reports?

For most organizations, yes, with one qualification. Stakeholders explore at their own pace, follow links between related elements and read current information rather than a snapshot from the last presentation cycle.

The qualification is repository quality. If your repository is not a reliable source of truth, Prolaborate shows that unreliability to the people you most want to impress, and an AI summary of a thin model is a thin summary. Effective adoption usually travels alongside a repository governance improvement program. Get the repository trustworthy first, then let the publishing and the AI do their jobs on top of it. That is the working method behind AI-Augmented Architecture.

Frequently asked questions

AI is only as good as the person prompting it, and that holds for the AI inside Prolaborate as much as anywhere. We can teach your architects to work this way on your own repository, or build and update the model for you so there is something worth publishing.