Contemporary design practices for industrial products often involve the detailed study of individual components and comprehensive analyses of the overall system. The latter is often performed by means of reduced-order approaches thanks to their versatility, ease of use, and a low computational effort. These tools are widely available as commercial software packages, with limited open and free alternatives, which are usually developed and tailored for very specific applications. In this work, a generalised object-oriented framework for the development and analysis of reduced order models is proposed, alongside its Python implementation. This toolbox is based on three main pillars: the use of graph theory for the representation of the computational domain, an advanced management system for units of measurement, and a generalised formulation for source and sink terms. The approach is presented alongside a number of case studies representative of physical problems of technical interest, including ventilation control in tunnels, and thermal management of electric vehicles.