Built for the DigitalOcean Gradient™ AI Hackathon

Mark4k is an intelligent assistant designed for students and researchers. It provides academic path advice, analyzes complex PDF documents (syllabi, assignments, textbooks), and generates complete project source code using the power of DigitalOcean Gradient AI (Qwen3 32B).

The project follows a standard Flask factory pattern for scalability and security:

mark4k/
├── app/
│   ├── static/
│   │   ├── css/
│   │   │   └── style.css      # Academic Dark-Gold UI Theme
│   │   └── js/
│   │       └── script.js     # Frontend logic & Smart ZIP/PDF handling
│   ├── templates/
│   │   ├── base.html         # Master layout
│   │   ├── chat.html         # Main Workspace (Chat/PDF/Code)
│   │   ├── login.html        # Authentication: Login
│   │   └── register.html     # Authentication: Register
│   ├── __init__.py           # App factory & Database initialization
│   ├── models.py             # SQLite User models (Flask-SQLAlchemy)
│   ├── routes.py             # Chat, PDF analysis, and File routes
│   └── utils.py              # AI API calls, PDF generation, & ZIP building
├── .env                      # API Keys & Secret configurations
├── requirements.txt          # Project dependencies
├── run.py                    # Entry point to launch the application
├── mark4K_previous.py        # [LEGACY] Includes SMTP Email Sending feature
└── README.md                 # Project documentation

The current modular application focuses on high-performance chat, PDF analysis, and secure project generation.

Note: The feature allowing users to draft and send academic emails via SMTP (Gmail App Passwords) is available in the mark4K_previous.py file located in the root directory. Users wishing to use the email module should refer to that specific script.

• Smart UI: Academic "Dark Mode" with deep navy and gold accents.
• Project Advisor: Multi-turn chat for academic guidance.
• PDF Analyzer: Upload documents to extract summaries and task requirements.
• Smart Code Generation: Automatically detects source code in responses and offers a Download ZIP button.
• Report Generation: Export any AI conversation as a branded PDF Document.
• Secure Auth: SQLite-backed user registration and login system.

Ensure you have Python 3.9+ installed.

pip install -r requirements.txt

Create or edit the .env file in the root directory and add your credentials:

SECRET_KEY=your_secret_flask_key
AGENT_URL=https://cjkgjyoo5oezhc77dzx2dmbx.agents.do-ai.run/api/v1/chat/completions
AGENT_KEY=your_digitalocean_agent_key

python run.py

The app will be available at http://localhost:5000.

• Backend: Flask, Flask-SQLAlchemy, Flask-Login
• AI Engine: DigitalOcean Gradient AI Agent (Qwen3 32B)
• Document Handling: pypdf (Extraction) & fpdf2 (Generation)
• Database: SQLite

Created for the DigitalOcean Gradient AI Hackathon 2025.