Optees — Optimization Toolkit
🟢 Active
Optees is an open-source toolkit for solving optimization and decision intelligence problems, designed to combine the power of Operations Research algorithms with a modern and accessible desktop interface. Written in Python and built according to a six-layer Clean Architecture, the project aims to offer a modular and scalable platform, suitable for both expert users and non-technical business figures. Current functionalities include linear programming (LP), mixed-integer linear programming (MILP), and Knapsack problems, but the goal is to extend the system towards heuristic methods, artificial intelligence, and more complex decision models, always maintaining a clear, guided, and transparent user experience.
Tech stack
Type
Category
Platform
Download
Checking...
Checking...
Checking...Do you like Optees and would you like to collaborate? Feel free to send me a pull request or open an issue!
Github
Algorithm Selection
From the homepage, the user can choose between different types of optimization problems: LP, MILP, and Knapsack. Each category has a descriptive preview and a consistent, yet specialized, interface. Advanced methods like Non-Linear Programming, Graph Theory, and AI are planned as future extensions.
Modular GUI • Multi-algorithm Entry Point • Future Expandability
LP Model Construction
The LP modeling interface allows simple addition of custom variables, constraints, and bounds. Fractional values, objective function offsets, and multiple constraints with different signs can be entered. The user has full control of the model thanks to a clear, responsive, and accessible GUI.
Model builder • Visual editing • Validated input
Result, Graph, and Feasible Regions
Optees visualizes the optimal solution to the problem, showing quantities, value, sub-totals, and objective. A 2D or 3D visualization (depending on variables) of the feasible region and optimal solutions is also generated. If multiple optimal solutions exist, the system will show the entire optimal line or plane rather than a single point.
Interactive Output • 2D/3D Graphs • Multiple Optimal Solutions
Modular 6-Layer Architecture
The project follows an advanced Clean Architecture with six separate layers: Utility, Core, Domain, Data, Application, and Presentation. This subdivision allows for testability, scalability, and long-term maintainability. Each layer has well-defined responsibilities, promoting separation of dependencies and independent module evolution.
Clean Architecture • 6 Layers • SOLID principles
Combine Operations Research and Artificial Intelligence
Are you interested in operations research and curious about how to apply it to your business context?