Introducing AI Capabilities in Prolaborate

FAQs

Prolaborate AI Capabilities

Enterprise Architect also provides AI capability. How does Kernaro AI differ from the AI in Prolaborate?

Kernaro extends AI across the wider Sparx ecosystem. It enables AI capabilities on top of EA through an embedded EA add-in, a centralised AI Hub for governance and shared access, and AI-powered capabilities that also work within Prolaborate. This allows teams to use AI not only inside EA, but also across collaborative architecture workflows, model analysis, documentation, validation, and enterprise-level governance.

Prolaborate AI didn't appear to require a client-side LLM, but to connect to Kernaro do clients need to have their own LLM installed?

As explained in the video AI capabilities in Prolaborate is powered by Kernaro AI and Yes — Kernaro requires an LLM configuration to enable AI capabilities.

Customers can configure their own cloud-subscribed LLM provider, such as Azure OpenAI, OpenAI, Anthropic, AWS Bedrock, or other supported providers, based on their enterprise preferences. Kernaro does not require the customer to install an LLM locally, but it does require access to a configured LLM provider for AI processing.

Does the summarise feature also address the accessibility challenge of diagrams for screen readers?

Yes, the summarize is indirectly benefiting user with accessibility constraints.

Is the AI feature opt-in / opt-out, or is it enabled by default for all Sparx EA users?

It is an optional selection who would like to subscribe for AI feature in Prolaborate can purchase Kernaro.

Can you specify which dashboard a widget should be placed on?

Yes, since we are in development stage we will certainly consider this request. Any charts or reports created from AI is allowed to be added to any existing dashboard specified by user or publish to a new dashboard.  

Kernaro

Where is Kernaro AI hosted?

Kernaro AI can be hosted either in the cloud or on-premises.

Is Kernaro AI SaaS only, or available on-premises as well?

Kernaro AI is available for both SaaS and on-premises customers. Feel free to reach out to sales@prolaborate.com for more information. 

Can Kernaro AI be used without any cloud connection?

If the question relates to working with a local EA model, then yes — Kernaro AI can be used with local EA repositories. If the question relates to the AI/LLM processing itself, Kernaro currently supports cloud-based LLM providers. Fully on-premises LLM support is not available at this time.

Where are the AI services and database hosted?

Kernaro AI is typically hosted on a separate server alongside the Prolaborate server to enable the centralised AI capabilities. It also includes an EA add-in version, allowing users to access AI-assisted features directly within Enterprise Architect.

Is Kernaro AI configurable in terms of which LLM it uses — or can a customer host their own on-premises model (if powerful enough)?

Cloud-subscribed LLMs are supported at present; on-premises LLMs are not supported at this time. Supported providers include:

  • Azure AI Foundry (OpenAI / Claude)
  • OpenAI
  • AWS Bedrock
  • Anthropic Claude

Is an outgoing internet connection required when using an on-premises deployment?

At present, only cloud-subscribed LLMs can be configured, so an outgoing internet connection is required for AI processing. On-premises LLM support is not available at this time.

Can Kernaro create document templates as well, for printed reporting?

Yes. Kernaro can be used to generate documents, including from existing document templates.

How Prolaborate Processing Model Information with AI

What happens if the imported data includes attributes that are not defined in the metamodel or MDG?

Import currently supports structured data handling based on the defined metamodel or MDG. If the incoming data includes attributes or stereotypes that are not yet defined, Kernaro AI helps identify those gaps and supports metamodel updates before the import proceeds.

Support for unstructured data handling through Kernaro Intelligence is currently in development and is planned for inclusion in the next major version.

What happens if elements already exist in the EA model — does Kernaro update them, or create duplicates?

If elements already exist in the EA model, Kernaro analyses the imported file and checks for possible duplicates based on the available context. Using Kernaro Intelligence, it highlights potential matches before the import, allowing the user to review and confirm whether existing elements should be updated or new elements should be created.

From an enterprise architecture governance perspective, does the solution validate Sparx EA models against ArchiMate metamodel rules and relationships?

Yes. Kernaro AI provides validation support for Sparx EA models against defined metamodel rules, including ArchiMate stereotypes, permitted relationships, and modelling constraints. Through Kernaro agents, model content can be analysed to identify rule violations, inconsistent relationships, missing attributes, and areas that require governance review or correction.

Can we create relationships between two different stereotypes via Excel import?

Yes. The data import shown in the webinar can include multiple stereotypes. Connectors can also be created between elements with different stereotypes, provided the source and target elements are correctly defined in the file.

Visit us