computerComputer Science (Year 3)
Where: 42Warsaw
Dates: 02.2024 — Present
Focused on systems architecture, low-level programming, and performance engineering. Courses include OS design, distributed systems, and advanced algorithms.
Architectural Status
I build systems: low-level architecture, performance tuning, and reliable distributed services. I enjoy pushing boundaries at the kernel and runtime layers, and designing maintainable infrastructure for ambitious AI and systems projects.
I focus on clean, maintainable code and observable systems that scale. Open to collaborations in systems engineering and ML infrastructure.
Where: 42Warsaw
Dates: 02.2024 — Present
Focused on systems architecture, low-level programming, and performance engineering. Courses include OS design, distributed systems, and advanced algorithms.
Where: WSKZ
Dates: 2024 — Present
Studying project management, organizational behaviour, and product strategy to bridge technical work with business impact and leadership.
Low-level architecture and high-performance language foundations.
Specialized AI runtimes and hardware-accelerated computation.
LLM_RUNTIMES
Ollama / vLLM / TensorRT
ACCELERATION
CUDA Optimization
CONCURRENCY
Asynchronous IO / Threads
Distributed systems, microservices, and containerization.
NETWORKING
TCP/UDP & Sockets
KUBERNETES
Cluster Mgmt / GitOps
VIRTUALIZATION
Docker / VFIO / QEMU
I manage my development environment as Infrastructure-as-Code. Utilizing a declarative NixOS stack, I focus on bridging software development with low-level kernel performance and hardware orchestration.
Semantic image search engine
An AI application enabling natural language search of image collections (semantic search) based on the Vision Transformer (SigLIP) model. I implemented embedding generation, a vector database (LanceDB) for fast searches, and performance optimization through multithreading and GPU acceleration (CUDA/Mixed Precision).
Local AI assistant
A local assistant controlling the system (brightness, volume, applications) with integration of LLaMA models via Ollama. The project includes command interpretation, TTS/STT, and automation of system operations. I gained experience with on-device AI models and the integration of multiple components.
A full-stack web application
A microservice-based application: backend in Node.js, frontend in TypeScript (Vite + Tailwind), reverse proxy in Nginx, containerization with Docker/Docker Compose, and a local database. Tested using Selenium in Python. The project taught me production architecture, service integration, and working with container infrastructure.
RFC 1459-compliant IRC server
A multi-user IRC server supporting channels, user roles, and socket-based communication and multiplexing mechanisms (poll/select). The project developed my skills in network programming, thread synchronization and working with low-level system APIs.
Raycasting-Based 3D Engine
A rendering engine inspired by Wolfenstein 3D: raycasting, textures, collisions, and rendering optimization. This project developed my skills in algorithmic thinking, computer graphics, and writing efficient C code.