At the moment, AI Assist can connect only to OpenAI models hosted on Azure Foundry.

No, Genie AI is language-agnostic. Queries and reports can be created in multiple languages, not just English.

AI Assist is a separate add-in for Sparx Systems Enterprise Architect. It works like any other add-in and must be installed independently.

Yes, AI features are available for fully on-premise installations.
AI Assist, Genie Hub, and Genie for Prolaborate can all be deployed in both on-premise and cloud environments.

Genie AI does not depend on a specific LLM. Organizations can connect their preferred LLM (enterprise or cloud-based).
Data is processed only during interaction and is not used for training or improving the LLM.

A default Jama MCP is not provided included. However, if Jama offers an MCP, Genie can integrate with it.

Pricing depends on customer requirements and available tiers.
For details please write us top kernaro@sparxsystems.com

The Genie server can be deployed either on-premise or in the cloud, based on customer preference.

AI Assist is a separate add-in and is not tied to a specific Enterprise Architect version bundle. It must be licensed and installed separately.

Full on-premise LLM support is currently on the product roadmap and is planned after the initial release.

Organizations can use their own LLM or select supported cloud-based LLM providers such as OpenAI or Anthropic Claude. Model data is processed only for the interaction and is not used for training or expanding the LLM’s knowledge base. Data usage policies depend on the agreement with the selected LLM provider.

A beta version is currently available. Commercial release is planned around mid-April.

Yes, AI Assist supports multiple modes:

• Plan Mode – Used for planning and outlining modeling activities.
• Ask to Edit Mode – AI proposes changes and requests confirmation before applying them.
• Auto Edit Mode – AI directly applies changes to the model.

The outcome depends on the instructions provided by the user. Using the available interaction modes such as Plan Mode or Ask to Edit Mode allows users to review proposed changes before they are applied to the model.

No. Genie is available for both cloud and on-premise deployments.

Yes. AI Assist can work with custom MDG technologies. It can be provided with instructions that make it aware of the model structure, and it also has the ability to read MDG definitions directly and incorporate custom attributes, stereotypes, and queries as part of its context.

Yes. Genie can leverage Enterprise Architect APIs to improve diagram layouts to a certain extent. The layout behavior can also be influenced or customized through user instructions.

In most cases, existing elements will be reused if they match the context. However, the behavior can be customized based on user instructions and the configuration provided during the interaction.

AI Assist is a separate add-in. Compatibility depends on the EA version, but newer versions are recommended.

A beta version is available now. The official release is planned around mid-April.

Yes. Pricing will be shared separately and will depend on licensing options.

AI Assist is currently available as an add-in for Sparx Systems Enterprise Architect and is not directly embedded within Prolaborate. To enable AI capabilities in Prolaborate, the Genie platform is required. As demonstrated in the webinar, Genie acts as the AI engine that connects to the architecture repository and enables AI-driven interactions, analysis, and automation within the Prolaborate environment.

Yes,AI Assist can also create specialized agents with defined instructions that can be linked to Enterprise Architect events.
These agents can automatically validate model content, enforce governance guidelines, and guide users during modeling activities. By attaching agents to specific EA events, organizations can introduce automated checks and intelligent recommendations that help ensure modeling standards and governance rules are consistently followed across the repository.