Assistant Setup
This section covers how to configure and manage AI models used by the AI Assistant in DecisionRules.
DecisionRules AI Assistant supports different AI integration options depending on your deployment type.
Cloud Version
In the Cloud version, AI integration is fully managed by DecisionRules.
No additional setup is required
The AI Assistant is available out of the box
Configuration is maintained by our team for optimal performance and reliability
About the Model and Limits
The assistant is powered by Gemini for Business. We are currently using the model gemini-3-flash-preview. The exact versions may change in the future.
Current usage limits for the assistant are defined by your subscription plan. If you are on a free trial, paid Lite or higher tier, you will have access to a limited number of AI generations each month.
Docker (On-Premise) Version
Make sure that you have set AI_ENGINE_URL environment variable. More details: Server (required)
In the Docker version, you can connect your own AI provider with the required credentials.
Configuration can be defined on the following levels:
Organization level
Environment level
AI Engine application level
Configuration Resolution
These configuration levels are evaluated in the following order:
Organization level
Environment level
AI Engine application level
This works as a fallback mechanism:
When an Organization-level configuration exists, it is used
If not, DecisionRules checks the Environment level
If that is also not configured, DecisionRules uses the AI Engine application level
Next Steps
For more details, see the dedicated pages for each configuration level:
Organization Level: Define AI Provider
Environment Level: AI assistant provider
AI Engine Application Level: DecisionRules AI Engine
Last updated
Was this helpful?

