MSc Data Science student at ELTE, focused on machine learning, research engineering, and AI safety.
I like building things that sit between research and engineering: reproducible experiments, data systems, evaluation pipelines, and tools for understanding real-world ML and AI systems.
Temporal-network audit of resource targeting during the DSEWiki autonomous-agent incident, developed for the Apart Research AI Incident Response Sprint 2026.
- reconstructed time-varying multi-writer activity from incident logs
- evaluated resource-targeting strategies under controlled graph interventions
- separated preregistered analysis from post-result exploratory controls
- produced a reproducible research paper with automated numerical claim verification
Python 路 Network Analysis 路 Research Engineering 路 Reproducible Research 路 AI Safety
24/7 data pipeline collecting Budapest GTFS-Realtime transit data for later delay prediction and ML research.
- fault-tolerant realtime collection at high frequency
- crash-safe Parquet storage and recovery
- data-quality manifests and integrity checks
- verified off-site backups and retention tooling
- Dockerized deployment on a Linux VPS
Python 路 Data Engineering 路 GTFS-Realtime 路 Parquet 路 Docker 路 Reliability Engineering
- PyTorch and deep learning
- machine learning foundations
- research engineering
- AI agents, evaluations, and technical AI safety
- building stronger end-to-end ML systems
E枚tv枚s Lor谩nd University (ELTE)
MSc Data Science
E枚tv枚s Lor谩nd University (ELTE)
BSc Computer Science
Python 路 PyTorch 路 Git 路 Linux 路 Docker 路 Pandas 路 SQL