AI Features Management

Decide Which LLM to Use

GermainUX AI supports multiple LLM integration approaches, allowing organizations to use their approved AI models and deployment standards.

Supported Integration Approaches

• Integration with models hosted internally and exposed through an OpenAI-compatible API endpoint

• Integration with externally hosted model runners such as Ollama

• Running models directly within the GermainUX AI Service using supported model files and frameworks such as Transformers

Information Required to Configure an LLM

To proceed, your organization can either configure the integration directly or provide GermainUX with the following information.

If Using a Hosted LLM API

• Hosted API endpoint URL

• Required authentication details or API keys

• Whether the LLM is exposed using an OpenAI-compatible API specification

If Deploying the LLM Locally

• Framework/runtime used by the model

• Model format provided

• Hardware or deployment requirements

• Any security or networking constraints

Once the hosting model and interface method are confirmed, GermainUX can configure the AI services accordingly and enable AI capabilities using your organization’s approved LLM.

Path: GermainUX Workspace > System > AI > Prompts > Model field > Select LLM

Online or Offline Execution

For most LLMs, GermainUX supports both:

• Online mode — Connect to an externally hosted LLM

• Offline mode — Run the LLM directly within the GermainUX AI Service

This provides flexibility for organizations with security, compliance, latency, or infrastructure requirements.

Path: GermainUX Workspace > System > AI > Prompts > Set Online/Offline for a prompt

Enable or Disable AI Features

By default, AI features are disabled to support security and compliance review processes.

Organizations can review, approve, enable, disable, or customize individual AI capabilities based on their internal policies and requirements.

Path: GermainUX Workspace > System > AI > Settings

Customize AI Prompts

AI prompts can be customized to leverage both preconfigured and custom KPIs, enabling organization-specific analysis, recommendations, and automation workflows.

Path: GermainUX Workspace > System > AI > Prompts > Select a prompt > Edit in right panel