AI Tutor is a web-based platform that supports professors and universities in providing students with automated and AI-supported exercises. The students explain the question from the exercise to the AI agent. The AI then helps the student to understand the topic by asking questions and providing hints. Once the student thinks that the exercise is solved, a second AI agent checks the conversation and gives feedback on whether the exercise was solved correctly or not. Tutors and admins have special roles to enable them to view the submitted conversations and give further feedback if necessary.
There are multiple configuration options available to customize the AI Tutor to your needs.
The privacy notice used on the website can be found in aitutor/pages/legal_infos/datenschutz.md. If you need to change it, you can edit this file.
A short version of this privacy notice is displayed on the register page. It can be modified in aitutor/pages/legal_infos/datenschutz_short.md
If you want to initialize the AI Tutor with some default information, you need to add a "config.toml" file to the root directory of the project. This file should contain information about defaultusers, prompts and other configurations. For more information about the config file and its content, please refer to the configfile documentation.
This project is managed with uv. See installation
instructions.
When you run any uv commmand for the first time, uv will automatically create a virtual environment in .venv at the package root.
You can activate this venv normally but for the uv commands, that's actually not needed as long as uv is called from the projects root directory. So instead of activating the environment, you can also run commands like this:
uv run <command>
Clone the repository:
git clone <URL-of-repository>
cd aitutor
Initialize database:
uv run reflex db init
Create local environment file .env with the following content:
OPENAI_API_KEY=<your key>(Optional) Add your own config.toml file for the initial configuration. See configfile documentation for more information.
To start the server in developer mode, run:
uv run reflex run
For production mode, please refer to docker.md for more information.
This repository contains a docker compose setup for easy deployment. More information on how to use this docker setup can be found in docker.md.
If you want to contribute to the project, please refer to the contributing documentation for developer infos.
All files in this package are provided under the AGPL v3 license, see LICENSE.
Copyright © 2024 Eberhard Karls Universität Tübingen, Distributed Intelligence/Autonomous Learning Group