Configure the AI Model Builder

The Semaphore Knowledge Model Management component includes an AI Model Builder tool that assists you in quickly building a model using an AI assistant. Through the AI Model builder, you can request suggestions for concepts to include in your model based on existing model data; then, add those concepts to your model with a couple clicks. The AI Model Builder requires that you provide your own Large Language Model (LLM) service and configure the tool to connect to your LLM. This topic describes how to configure the AI Model Builder to use your LLM.

Note: To use the AI Model Builder, you must provide access details of an LLM to which you have a subscription. Progress Software Corporation is not a party to your agreement with your provider and has no control over the service or its availability. Progress cannot be held liable for any output, errors, damages, losses, or other results from your use of the integrated third-party AI. Be aware that many LLM providers charge costs for using their service.

Related videos

In this video, you learn how to enrich your semantic model in Semaphore using the AI Model Builder. This includes learning how to use the AI Model builder for alternative labels, narrower concepts and advanced suggestions.

Before you begin

Before you begin, you need the following:

• Knowledge of how to connect to your LLM Model Provider.
• You must be assigned the SemaphoreSuperAdministrator role.

Configure your AI Model Builder

To configure the AI Model builder:

1. Login to Semaphore Studio with a user assigned the SemaphoreSuperAdministrator role.

2. Select the Services tab.

3. Select the AI Model Builder from the side menu.

4. From the LLM Model Provider field, select your Large Language Model provider from the list of supported providers. For example, Azure OpenAI. The configuration fields that are specific to your LLM Model provider are exposed.

5. Provide values for the configuration fields for your LLM Model Provider. For example, for Azure OpenAI, provide the values for the following fields:

   • API Key: Specify the API key that is used to authenticate to your Azure OpenAI.
   • Azure OpenAI Endpoint: Specify the endpoint for your Azure OpenAI LLM service.
   • Azure Deployment Name: Enter the version of your LLM. For example, GPT4-32k.

   Notes:

   • You can find configuration information for your Azure OpenAI in your Azure Portal.
   • If you have restricted access to your Azure OpenAI instance, ensure that the apparent IP address of your server is whitelisted in the Azure portal. One method of finding this IP address by running the following command on your server:
     Copy

     curl https://ipecho.net/plain
     

6. Click Test to test your configuration information. You will receive a confirmation message if your information is valid.

7. Click Submit to save and apply your configuration.

See also

• How to add a concept with the AI Model Builder