Skip to main content

Configure AI Provider

Learn how to add your own AI provider to UnderControl for AI chat and smart features like task summaries and intelligent categorization.

Prerequisites​

  • An UnderControl account (registered user, not visitor)
  • An API key from an AI provider (OpenAI, GitHub Models, or any OpenAI-compatible service)

Steps​

Step 1: Go to Settings​

Click your username at the bottom-left corner of the sidebar to open the Settings page.

Step 1: Go to Settings

Step 2: Open AI Configuration​

In the settings sidebar, click the AI Configuration section.

Step 2: Open AI Configuration

Step 3: Add an AI Provider​

Click the + Add Provider button. In the dialog, fill in:

  • Name: A friendly name for the provider (e.g., "Github")
  • Provider Type: Select the provider type (OpenAI, Custom OpenAI-compatible, etc.)
  • API Key: Your API key from the provider
  • API Base URL: The provider's API endpoint (e.g., https://models.github.ai/inference)
  • Model: The model to use (e.g., gpt-4o)

Step 3: Add AI Provider

Step 4: Test the Connection​

Click the Test Connection button to verify your API key and configuration are correct. A green checkmark confirms the connection is successful.

Step 4: Test Connection

Step 5: Set Availability​

After creating the provider, configure where it can be used:

  • Backend (toggle): Enable for server-side AI features such as task summaries and smart categorization
  • Frontend (toggle): Enable for browser-side AI chat

Both toggles will show Available when enabled.

Step 5: Set Availability

Step 6: Chat with AI​

Open any task and click the AI Chat panel on the right side. Select your newly configured provider from the dropdown menu to start a conversation.

Step 6: Chat with AI

Result​

You now have a fully configured AI provider. You can:

  • Use AI chat from any task's detail page
  • Get AI-powered task summaries and suggestions (if backend is enabled)
  • Switch between multiple providers in the AI chat dropdown

Tips​

  • You can add multiple providers and switch between them in the AI chat panel
  • Use Custom (OpenAI-compatible) provider type for services like GitHub Models, Azure OpenAI, or local LLMs that support the OpenAI API format
  • If you only need chat features, enabling just the Frontend toggle is sufficient
  • Backend AI features require the provider to be enabled on the backend toggle