Yassine Chibani

– Completed projects in C and C++ following 42 School's pedagogy: no formal courses, self-taught language and programming concepts, peer-to-peer learning for project reviews to validate skills, with code adhering to strict development norms. – Mastering low-level languages like C and C++ provides rock-solid programming foundations, making it easier to learn other languages. – The common core covers a wide range of computer science concepts through diverse projects: – Recreating standard C library functions to deeply understand them. – Developing a bash-like shell in C (managing multiple processes). – Network & System Administration: Docker project setting up an nginx-php-mysql server (using docker-compose), plus VM configuration (virtual disk partitioning, SSH and firewall setup, password policy, bash scripts, etc.). – Implementing a sorting algorithm. – Building a multi-threading program. – Creating a 3D maze game (inspired by Wolfenstein 3D) using raycasting for 3D rendering with the libX library. – Use of C++ containers (stack, vector, map). – Building a full-stack website that implement's a pong game, real-time chat within channels, 2FA and login authentification

– Developed a comprehensive LLM-powered MVP that transforms natural language questions into accurate SQL queries, streamlining data access for business users across the organization. – Conducted in-depth research and benchmarking of leading Text2SQL frameworks including Vanna.ai and WrenAI, while establishing seamless data source integrations with DuckDB databases to ensure real-time query execution against production datasets. – Designed and deployed a specialized chatbot interface tailored for handling event-related user inquiries, providing instant, context-aware responses powered by fine-tuned language models to enhance user experience during live operations. – Architected an enterprise-grade Ollama Proxy that intelligently routes Python class calls from local Ollama instances to Azure private cloud-hosted models, enabling secure, scalable LLM inference while maintaining full compatibility with existing development workflows and ensuring data privacy compliance.