I am a Junior Software Engineer (Python / C) currently building robust, low-level technical expertise at the 42 Prague Coding Institute. I bring a unique advantage to development teams: 25+ years of experience as a translation agency founder and a project manager. Today, I apply that same uncompromising QA, system-level thinking, and execution discipline to writing clean, PEP 8-compliant code.
Opportunities as a Python / C developer to build robust systems and write clean, PEP 8-compliant code. I am particularly interested in backend development, Retrieval-Augmented Generation (RAG), and AI cost-optimization. I bring a unique blend of 42 Prague engineering rigor and decades of entrepreneurial project management experience.
-
Backend Architecture & Optimization: Designing robust Python/C systems with a focus on memory management and resource efficiency.
-
AI Implementation: Building RAG pipelines and deterministic LLM function-calling engines (constrained decoding).
-
Technical Quality Assurance: Rigorous testing, static typing (MyPy), strict linting (Flake8), and memory leak prevention (Valgrind).
-
Engineering Execution: Leveraging 25 years of business management experience to ensure accurate requirements scoping, risk mitigation, and flawless project delivery.
Languages: Python (3.13), C, PHP, JavaScript, Bash/Shell
AI & Data Science: Retrieval-Augmented Generation (RAG), Constrained Decoding, NumPy, HuggingFace Transformers, PyTorch, LLM integration
Systems & Concurrency: POSIX Threads (pthreads), Mutexes, Memory Management (Valgrind), Custom Priority Queues (FIFO/EDF)
Networking & Infrastructure: TCP/IP, IPv4/IPv6 Subnetting, OSI Model, Routing, Virtual Machines, Docker
Tools & Quality Assurance: Git, Pytest, Flake8, MyPy, Valgrind, Makefiles
Libraries & Frameworks: Pygame, MiniLibX, Pydantic, FastAPI (if applicable)
- Core Programming: advanced C/Python programming, memory management, variadic functions, concurrency, asynchronism, constrained function calling, file I/O
- Python Development: Object-oriented programming, decorators, comprehensions, design patterns, Pygame, MLX
- Algorithms & Data Structures: Graph theory, pathfinding, sorting algorithms, linked list manipulation
- System Administration: Virtual machine deployment, LLMP stack, Debian/Linux, Network Configuration, Subnetting
RAG against the machine - A custom Retrieval-Augmented Generation (RAG) engine built in Python, implementing code-aware tokenization, multi-format chunking, and inverted indexing. It leverages BM25 search scoring to optimize context retrieval and efficiently surface highly relevant document fragments for LLM integration. Tech Stack: Python, BM25, Inverted Indexing, Tokenization, Chunking Algorithms.
Call_me_maybe - A deterministic LLM function-calling engine, implementing constrained decoding with NumPy-based logit masking to force strictly validated JSON output from LLMs, bypassing probabilistic randomness with 99% structural success rates. Tech Stack: Python, NumPy, Hugging Face Transformers, FSM-based Logit Injection.
Pac-Man - Complete, playable Pac-Man game in Python, using OOP, MiniLibX/Pygame graphical libraries and a modular, reusable architecture. Tech Stack: Python 3.13, Pygame, MLX, Back-end/Front-end, Github, Project Management.
Get-IPv6 - IPv6 subnetting and routing, hexadecimal notation, prefix lengths, subnet calculations, hierarchical subnetting, address types, scopes, types and common IPv6 networking concepts. Tech Stack: IPv6 subnetting, CIDR/prefix notation, hexadecimal and binary calculations.
NetPractice - Application of practical network configuration, requiring the calculation and assignment of IP addresses, subnet masks, and routing tables to make non-functioning network topologies operational. Tech Stack: OSI layers (Network Layer), subnetting, CIDR, default gateways, routing tables and switches.
Codexion - Advanced concurrency simulation featuring POSIX threads with custom priority queue scheduling (FIFO/EDF), mutex-based synchronization, and deadlock prevention. Demonstrates mastery of thread safety, condition variables, resource arbitration and the Dining Philosophers problem variant.
Fly-in - Efficient drone routing system that navigates multiple drones through connected zones while minimizing simulation turns and handling movement constraints (Cooperative Space-Time A* algorithm, Pygame, heapq(), frozenset())
A-Maze-ing - Python CLI maze generator and solver using BFS/DFS algorithms with interactive gameplay and terminal animations
Push_swap - Complex algorithmic challenge: sorting integers in linked list stacks with a limited instruction set
Python - Advanced Python modules covering OOP, decorators, inheritance, polymorphism, and design patterns
Libft - Custom C library replicating standard libc functions with utility additions
Printf - Full implementation of printf() with format specifier support
Get_next_line - File reading function demonstrating buffer management and static variables
Born2beroot - Linux system administration project: VM deployment with LLMP stack and WordPress
Python-Piscine - Intensive introductory course building solid Python 3 fundamentals
Learn more: GitHub Repositories | 42 Prague

