Set up your institution, step by step. Start here.

How to Configure AI and LLM Integrations

Step-by-step instructions to connect your platform to external AI providers and manage student and administrator chat assistants.


6. AI & LLM (AI models for intelligent features)

Before you can use the AI Chat assistant, you must configure at least one AI provider.

The OpenAI Integration dialog, showing the pattern shared by every AI provider

Every card in this category — OpenAI, Anthropic Claude, Google Gemini, Groq, and Custom Provider — opens the same dialog, just pre-filled for that provider:

  1. Click Configure (or Set up) on the card.
  2. Toggle Enable [Provider] Integration on.
  3. Enter your API Key — get it from the linked console (e.g. OpenAI Console). OpenAI also shows an optional Organization ID, needed only if you belong to multiple OpenAI organizations.
  4. Click Test & Fetch Models in the Test Connection box. It calls the provider's live model list and shows one of three results: a green box listing the models available (and a reminder to pick one in AI Chat Settings), an amber box if the key works but the provider doesn't publish a model list (you'll type the Model ID manually later), or a red box with the connection error.
  5. Read the 🔐 Security Note — API keys are encrypted with AES-256-GCM before storage — and the 💡 Usage Tips (monitor usage in the provider's console, set spending limits, use separate keys for dev/production).
  6. Click Save Changes.

The AI & LLM category: OpenAI, Anthropic Claude, Google Gemini, Groq, Custom Provider, and AI Chat Settings

An earlier version of this guide listed a "Voyage AI" card that does not exist, and omitted the real fifth provider, Custom Provider. Custom Provider connects any OpenAI-compatible API — OpenRouter, DeepSeek, Together, Fireworks, Ollama, LM Studio, and more — and its dialog adds two extra fields before the API Key: a Provider Name (a friendly label, e.g. "OpenRouter") and a Base URL (the API endpoint, e.g. https://openrouter.ai/api/v1 or http://localhost:11434/v1 for a local Ollama server).

AI Chat Settings (You must enable at least one provider above first — otherwise this dialog just shows "No AI Providers Configured".)

  1. Find the AI Chat Settings card and click Configure.
  2. The dialog has one tab, Student AI — configure whether students can chat with an AI assistant about their enrolled courses, payments, exam results, and assignments.
  3. Toggle Enable AI Chat for Students, pick an AI Provider from the ones you enabled, and choose a Model. The model list is fetched live from that provider; if none of the curated models match what came back, pick from the live list, or switch to manual entry to type an exact Model ID yourself.
  4. Click Save Changes.

There is no separate "Admin AI" tab. An earlier version of this guide described one — it doesn't exist in the interface. Only student-facing AI chat is configured here.

Related: Managing Integrations

#ai #llm #openai #anthropic claude #google gemini #groq #voyage ai #chat settings #integrations