I design and build software systems, tools and prototypes — from desktop applications to backend services and interactive concepts.

• Paper-CPU-Von-Neumann-Architecture-Simulator

ISA Simulation (2020): A C-based virtual CPU implementing a custom 8-bit Von Neumann architecture. Features a modular execution engine and an interactive stepper/debugger.

• MyShelf-A-custom-Unix-Shell

System Architecture: C-based Unix Shell demonstrating core OS concepts, originally built in 2019 and refactored for memory stability in 2026.

• Text-Adventure-DSL-Interpreter

Custom DSL & NLP Engine (2022/2026): A multi-layered SLY-based pipeline (Lexer/Parser) interpreting natural language commands. Features a domain-agnostic core and lazy evaluation for complex state management.

• SeaWar-Terminal-Tactical-Battle

Game Engine & AI (2021/2026): A Java-based naval tactical engine implementing OOP Design Patterns and MVC principles. Features a multi-state heuristic AI (Hunt & Target logic) and a custom TUI with a retro scientific calculator aesthetic.

• Logic-Programming-Inference-Engines-Prolog

Logic & Optimization (2020 / 2026): A showcase of the declarative paradigm using SWI-Prolog. Features an interactive rule-based expert system and a Magic Square solver optimized with Constraint Logic Programming (CLP:FD) to handle high-dimensional search spaces.

• Custom-HTTP-Stack-Symbolic-NLP-Middleware

Symbolic AI & NLP (Licence/Bachelor Legacy 2020): A from-scratch implementation of the ELIZA chatbot engine and a custom HTTP stack in Common Lisp. This project explores the roots of Conversational AI by bridging stateless web protocols with stateful symbolic logic.

• SOM-Kohonen-C-Engine

From-Scratch AI (2022/2026): A high-performance Self-Organizing Map implementation in C99. Memory-safe (Valgrind verified) and optimized for high-dimensional clustering on datasets like Fisher's Iris and Palmer Penguins.

• Deep-RL-Sim-Environments

Research Synthesis (2022): Exploration of DRL taxonomies and Sim-to-Real challenges.

• UScrap2Gather

Data Pipeline (2022/2026): A robust Python/PySide6 tool for real-time Twitter scraping. Features multi-threaded collection, PostgreSQL integration, and a sophisticated GUI for campaign management.

• Hate-Speech-Detection-Analysis 🚧 [Refactoring in Progress]

Hybrid ML & Transformer-based Analysis (2022): Research project on French hate-speech detection and dataset creation. Implemented a sophisticated multi-stage pipeline:

• Data Engineering: Large-scale scraping followed by POS Tagging & Shallow Parsing to extract syntactic structures.
• Unsupervised Mining: Applied LOF (Local Outlier Factor) and Random Forest for anomaly detection, using K-Means to cluster lexical patterns without prior labeling.
• Deep Learning Inference:Integrated Transformer models (NLP) to resolve context-dependent semantics, significantly improving detection accuracy over traditional keyword-based heuristics.

• 👀 I’m interested in software engineering, architecture, automation, data workflows, and interactive systems/gaming concepts.
• 🌱 Current Focus: Modernizing legacy systems, cloud-native architecture, and Procedural Content Generation (PCG).
• 💾 I progressively publish selected components, utilities and experimental projects here.