Set Up AI Integration
Configure an AI API key and enable AI features for your tenants.
Before you start
| Who this is for | CAST Imaging administrator |
| Role required | Admin or Super Admin |
| Prerequisites | An OpenAI or Azure OpenAI API key; internet access to the AI provider (or a proxy) |
Step 1: Get an API key
OpenAI: Create an account at openai.com, go to View API keys, and click Create new secret key.
Azure OpenAI: You need an Azure subscription with access to Azure OpenAI. Deploy a GPT-4 model. Collect the API key, endpoint URL, and deployment name.
Model requirement
The AI chatbot feature requires GPT-4. The Explain features work with GPT-3.5 Turbo or GPT-4.
Step 2: Enter the API key in the Admin Center
- Open the Admin Center.
- Go to AI Settings (or License management in older versions).
- Select OpenAI or Azure OpenAI.
- Enter the API key (and for Azure: the endpoint URL and deployment name).
- Click Update.
[TODO: Screenshot of the AI Settings panel]
Step 3: Enable AI features per tenant
AI features are disabled by default. Enable them for each tenant where you want users to have access.
- In the Admin Center, go to Applications.
- Select the tenant.
- Toggle AI features to on.
[TODO: Screenshot]
Verify the setup
Ask a user in the tenant to right-click an object at Objects level. They should see Explain with AI in the context menu.
If the option is not visible, see AI features are unavailable or return errors.
Next steps
Feedback
Was this page helpful?