AI Tutor

This project provides an application for automatic correction and analysis using AI model The service takes an essay and a predefined question as input and returns detailed feedback in HTML format.

Features

• Essay evaluation with AI-powered feedback
• HTML-formatted response suitable for integration into applications
• Docker containerization

Prerequisites

• Docker and Docker Compose installed
• Hugging Face API key

Testing

For testing purposes, you can:

1. Clone the repository:

git clone (https://github.com/Pleias/BSF_AI_tutor.git)
cd BSF_AI_tutor

2. Write you hugging face key into .env file

HF_TOKEN=your_hugging_face_key_here

3. Run Docker Compose

docker-compose up --build

4. Open http://127.0.0.1:8090 in your browser Try the interface to see example inputs and outputs

Integration

Available endpoints:

1. GET '/' - renders test page with form
2. POST '/submit' - handles essay evaluation

Integration Example

import requests

# For essay evaluation
url = "http://127.0.0.1:8090/submit"
data = {
    "essayInput": "User input text",
    "questionInput": "Question for the writing exercise"
}

# Make API call and get HTML response
response = requests.post(url, data=data)
html_feedback = response.text