Personal Projects
A collection of projects I've worked on, ranging from university theses to personal experiments and open-source contributions.
Distributed Version Control System
A custom Distributed Version Control System (DVCS) developed as an engineering thesis. Features a CLI, a desktop GUI, and an SSH server with role-based access. The architecture includes a language-agnostic plugin system utilizing child processes and JSON contexts.
Group Expense Sharing App
A comprehensive full-stack mobile application for managing and optimizing shared group expenses. It incorporates a smart debt simplification algorithm to minimize transactions. The backend features a robust RESTful API built with ASP.NET Core, utilizing CQRS with MediatR, and Entity Framework Core for data persistence.
Car Rental Aggregator & Comparison Platform
A comprehensive cloud-native car rental platform featuring a price comparison engine and provider integration APIs. The backend is built with ASP.NET Core utilizing CQRS, MediatR, and Entity Framework Core, deployed on a robust Microsoft Azure architecture. The frontend is a Blazor WebAssembly application featuring Google Maps integration and Microsoft Entra ID for secure, role-based access.
Search Algorithms in the Breakthrough Board Game
A high-performance implementation of the Breakthrough board game featuring advanced AI agents. The project explores Monte Carlo Tree Search (MCTS) modifications, evaluating them against a Minimax baseline with Alpha-Beta pruning. The computational layer and interactive GUI are built entirely in Rust, while Python is used for data analysis and evaluation of agent performance.
Learn More
Ant Colony Optimization Analysis
An implementation of Ant Colony Optimization algorithms and an analysis suite for the Distance-Constrained Capacitated Vehicle Routing Problem (DCVRP). The high-performance computation layer is built in Rust, providing CLI executables for Greedy, Classic, MAX-MIN, and Rank-Based Ant Systems. The experimental layer, which is handled in Python, evaluates algorithm convergence and stability.
Learn More
Genetic Algorithms Sudoku Solver
A WPF desktop application that solves 9x9 Sudoku puzzles using genetic algorithms instead of traditional backtracking methods. The project features a highly modular architecture that allows dynamic customization of representation, fitness functions, selection, crossover, and mutation. It includes real-time fitness charts and execution analytics to visualize the evolutionary process.