QLX AI Powered ML System is a demo web application that uses machine learning to analyze and predict data from a database.

Download the release from here and run the installer.

The release version includes a settings page where you can configure the LLM provider, API URL, model, database URL, and API key.

Also, don't forget to shutdown the application to abort the background process.

• Python 3.11.6+
• Node.js 24.11.1+

Open a terminal and run the backend:

cd backend # Create and activate virtual environment python -m venv venv # Windows: .\venv\Scripts\activate # Linux/Mac: # source venv/bin/activate # Install dependencies pip install -r requirements.txt # Set environment variables cp .env.example .env # Run server uvicorn app.main:app --reload

Open another terminal:

cd frontend # Install dependencies (if not done) npm install --legacy-peer-deps # Run dev server npm run dev

sqlite:///../example.db sqlite:///../example_casino.db

Note: on the .exe version, the relative path are: sqlite:///../../example.db sqlite:///../../example_casino.db

Note: Try to use absolute path on the .exe version.

hana://user:password@host:port

• Ensure hdbcli and sqlalchemy-hana are installed (included in requirements.txt).
• The port is typically 3xx15 where xx is the instance number (e.g., 30015 for instance 00).

postgresql://user:password@host:port/dbname

• For AWS RDS, the host is usually the endpoint provided in the AWS Console (e.g., mydb.cabc123.us-east-1.rds.amazonaws.com).

mysql+pymysql://user:password@host:port/dbname

• For AWS RDS, ensure the Security Group allows inbound traffic on port 3306 from your IP.

1. Open Browser to http://localhost:5173.
2. In the "Database Connection String" field, enter a valid string (see examples above). (Note: example.db was created in the root directory. The path is relative to where the backend runs. Since we run from backend/, we use ../example.db. Alternatively, use absolute path).
3. Click "Analyze Database".
4. Monitor the "Pipeline Results" page:
   • Adaptation: Check if Python code appears in the code block.
   • Execution: Check if "Model Performance" metrics (MSE, R2) appear.
   • Insights: Check if a text summary appears.

• Connection Error: Ensure the path to example.db is correct relative to the running backend process, or use absolute path.
• Missing Dependencies: Check requirements.txt for scikit-learn (it might be missing).

scikit-learn was not explicitly added to backend/requirements.txt but is needed by the generated code. The execution environment needs it.

A sample SQLite database is created in the root directory for testing.

Table: houses

Use connection string: sqlite:///../example.db (if running from backend/ dir) or auto-resolved path in Dashboard.