Skip to main content

Set Preferences

If you have worked in the application for some time, you will probably have developed a preference for certain settings. You can configure these preferences on the Preferences page.

Enable Microsoft integrations personally​

Has the administrator approved the Microsoft integrations for your environment? Open profile → Preferences → Tools → Connections, enable the desired Outlook, Teams, or SharePoint connection, and complete the Microsoft sign-in if needed. Signing in to the app with Microsoft does not automatically enable these connections.

Your administrator manages the app and permissions; you do not enter a client ID or secret and do not choose separate write permissions. Read Enable Microsoft integrations for all steps, example commands, and troubleshooting help.

Where to find it​

You can find the preferences in the top-right corner of the application, in the profile menu, under Preferences.

Preferences in menu

The Preferences page​

Below is a screenshot of the Preferences page.

Set Preferences

Drive connections​

Under Drive connections, you can connect cloud storage to AI-School. After that, you can upload files from File Manager or from a document preview directly to a selected cloud folder.

You can set up the following types of connections:

  • Google Drive: choose a folder from your own Google Drive.
  • Shared Drive: choose a folder from a shared Google Drive.
  • OneDrive: choose a folder from your own OneDrive.
  • SharePoint: search for a SharePoint site, choose a document library, and then choose the desired folder.

Adding a connection works as follows:

  1. Open Preferences.
  2. Go to Drive connections.
  3. Click Add connection.
  4. Choose the type of cloud storage.
  5. Click the folder button to choose a target folder.
  6. Save the connection.

For Google Drive and Shared Drive, you must be signed in with a Google account or grant permission again. For OneDrive and SharePoint, you must be signed in with a Microsoft account or grant permission again. With SharePoint, you do not need to enter site, library, or folder IDs manually; you choose everything through the folder picker.

You can create multiple connections. When uploading, you then choose which connected folder the file should be sent to. Set a connection to inactive if you do not want to use it temporarily.

Migrate data​

If the administrator has enabled migration for your environment, you will see the Migration tab in your account. This allows you to transfer data from the configured source environment, for example Demo or Schoolplein.

First click Check available sources. AI-School will then check whether there is an account in the source environment that belongs to you. If data is found, you can use Start migration to choose what you want to transfer.

You can choose from items such as:

  • profile data;
  • chats and chat messages;
  • files.
  • assistants you created yourself, including linked files and images (existing sharing settings are not transferred).

Staff and student data are not migrated. The migration runs in the background. You can keep using the application while the migration is running.

After a migration has completed, you cannot start the same migration again. This prevents the same data from being transferred twice.

Default AI provider​

Choose which AI provider you want to use by default. The environment determines which current language and image models belong to this provider. This allows the administrator to update models in the future while you keep the same provider as your preference.

You will see a short explanation for each provider and the models currently available. Click a model name to view more information about that model.

If your administrator offers both the cost-effective and the higher-quality model of a provider, you also choose which of the two new chats use by default. You choose the top model per chat.

Available providers may include, for example:

  • OpenAI: broad language, reasoning, and image models.
  • Claude: careful answers, long texts, and documents.
  • Google: Gemini models with Google Search and image generation.
  • European AI: European providers, with Mistral for text and FLUX for images.

Not every environment always has all providers available. If a provider has been disabled by the administrator, it will not be shown.

Speech support and text to audio​

If you need speech support, you can set this option to "Yes". An extra button will then appear next to the answer. This lets you have the text read aloud.

For Text to audio, you can also save personal settings from the text-to-audio input field. AI-School will then remember your chosen model, language, voice, and system prompt. The available TTS models and voices come from the AI provider available in the environment. For example, if you choose Google as your preferred provider, the linked Google speech models will also be used as the default where possible.

You can always restore saved text-to-audio settings to the default settings.

Image generation preferences​

You can also save personal settings for Text to image. AI-School will then remember, among other things:

  • the chosen image model;
  • the default format (Automatic, Square, Landscape, or Portrait);
  • style;
  • number of variants;
  • optional color preferences.

With Restore defaults, you remove your saved image settings. In edit mode, AI-School does not choose a manual format; the model then keeps the source image dimensions as much as possible.

File processing notifications​

Receive email notifications when files have been processed. If this option is enabled, you will receive an email when a file has been successfully processed or if an error has occurred.

Language preference​

Select the language you want to use the application in. The application supports, among others:

  • 🇳🇱 Nederlands
  • 🇬🇧 English
  • 🇫🇷 Français
  • 🇩🇪 Deutsch
  • 🇪🇸 Español
  • 🇮🇹 Italiano
  • 🇵🇹 Português
  • 🇵🇱 Polski
  • 🇷🇴 Română
  • 🇩🇰 Dansk
  • 🇳🇴 Norsk
  • 🇸🇪 Svenska
  • العربية
  • हिन्दी
  • 日本語
  • 中文
  • Русский
  • Türkçe
  • Українська
  • فارسی

Voice chatbots​

Manage which default voice chatbots are available on your dashboard. You can choose which languages you want to use for voice chats:

  • 🇳🇱 Nederlands
  • 🇬🇧 English
  • 🇩🇪 Deutsch
  • 🇪🇸 Español
  • 🇫🇷 Français

Tools settings​

The environment determines which tools are available. In Preferences → Tools, you then choose which visible tools and connections may be used by default in your chats.

For regular tools, you can choose a personal setting:

  • Automatic: the AI decides itself when the tool is used.
  • On: the tool is always directly available in the chat.
  • Off: the tool is not available to you.

A personal setting never makes a tool visible if the administrator has hidden it. With Restore default values, you use the environment settings again.

Available tools​

  • Create image: Generate images based on descriptions
  • Create practice questions: Automatically create practice questions about a topic
  • Search the internet: Search for current information on the internet
  • Search images: Search for images on the internet
  • Get weather: Request the weather for a location
  • Check AI content: Check whether text was generated by AI
  • Search Wikipedia: Search Wikipedia for encyclopedic information
  • Google Routes: Plan routes and get directions
  • Google Places: Find information about locations and businesses
  • Wolfram Alpha: Perform complex calculations and data analysis
  • Wikiwijs lessons: Search Wikiwijs for educational lesson materials (only available in school environment)

Tip: Click the information button in Tools settings for more explanation of how these tools work.

Connections​

Under Connections you will find the Google and Microsoft connectors and MCP servers that are available to you. Enable a connection once and complete the sign-in if needed. The choice is saved in your account and therefore also applies in new chats and on other devices.

Your organization reviews the Google and Microsoft connections under Management → Tools → MCP. The detailed explanation of data usage appears only after you enable the relevant connection. The screen also shows the MCP access for the environment. Is MCP access turned off when you enable a connection? In the warning, choose to enable MCP and continue. This turns on MCP for staff. When connecting your own account, you will see a short message about the data and AI processing, with a link to the privacy statement. Continuing starts authorization at Google or Microsoft; canceling stops the connection.

  • Built-in Google and Microsoft connections work on web, mobile, and desktop.
  • Remote Streamable HTTP MCP servers work on all three platforms.
  • Local stdio MCP servers are only available on the computer where you install them in the desktop app.
  • The chat settings show which connections are available in the current chat.

The built-in Google and Microsoft tools are provider-independent. Your preferred provider therefore does not change when you enable a connection. The connections remain available with OpenAI, Claude, Gemini, and Mistral, as far as the selected model supports tool use.

On desktop, you use the same Google and Microsoft connections as on web and mobile. Local stdio MCP servers, for example for Nextcloud, are separate from these account connections and apply only on that computer.

Read more at Connections and MCP.

General instructions and information​

In this section, you can set the so-called "system prompt".

With a system prompt, you can give instructions to the AI model. It is an instruction you can provide to the model to shape the task or behavior of the model specifically. You can specify what kind of output you expect from the AI model. It is a way to guide the model and prevent it from giving irrelevant or unwanted answers.

Some examples of system prompts​

I want you to be a friendly and helpful assistant."

"Always remain professional and respectful in your language use."

"Focus on giving concrete, to-the-point answers without unnecessary details."

The system prompt, as it were, sets the tone and expectations for the specific conversation or task you want to perform with the AI model.

You can also come up with more advanced system prompts for specific tasks, such as summarizing a text: "Carefully analyze the given text and provide a concise but informative summary of the main points. Identify the core message and most important details."

System prompt and the various AI models​

System prompts do not have the same effect on the outcome for every AI model. Large models usually follow instructions more carefully than fast or cost-effective models. Interesting to try out, then, and see which provider and which model works best for your input.

More information​

In the input window you will find much more explanation about system prompts. Click the “information” button for that. See below for a screenshot of the information window.

Information window System prompts

On this page