Selected works written in MDX. Explore project documentation, tech specs, and architectural details.
A restaurant management platform connecting guests, waiters, kitchens, and owners through QR ordering, live kitchen displays, floor plans, and tenant-aware operations.
A mobile-first fitness tracker for strength workouts, runs, and rides, with editable training plans, exercise history, progress analytics, and offline workout logging.
A daily guessing game about AI models, with Classic, Emoji, and Timeline modes, a React frontend, a Rust/Axum API, and server-authoritative progress backed by SQLite.
A computer vision project for reading physical tabletop RPG dice from a camera, managing multi-stage rolls, correcting uncertain detections, and sending confirmed results into Foundry VTT.
An AI-assisted portfolio intelligence and controlled trading system that combines position monitoring, reporting, opportunity scans, and strict safety layers before any trade automation is allowed.
A movie recommendation app built for an Advanced Databases exercise, using SurrealDB as a hybrid document and graph database after the project shifted away from the original Neo4j idea.
Retrieval-augmented assistant for Polish legal codes that lets users upload PDF acts, index them into Qdrant, and ask grounded questions through a FastAPI backend and Svelte admin UI.
A cow-vs-buffalo object detection pipeline with Faster R-CNN, PyTorch Lightning, Albumentations, and COCO-style evaluation, completed as a semi-project during the transition toward AI Engineering.
A League of Legends tournament and internal workplace stream setup for over 40 players, spectators, multi-stage play, rank-aware team selection, LAN finals, live camera feeds, OBS scenes, and commercial breaks.
A URL shortener with a historical Azure deployment pipeline that used Pulumi, GitHub Actions, Docker, Azure Container Registry, App Service, and Cosmos DB before the project moved to a Cloudflare Workers deployment.
A small university project for an Introduction to Artificial Intelligence class, where we used NEAT and Pygame to train Pong paddles over several generations.