HuSpaCy is a hobby project of mine: it contains Hungarian models for spaCy and integrates well with Hungarian NLP resources. The project got academic support in 2021, and it has grown to be a fully pledged industrial-strength NLP pipeline which is now widely used in both commercial and academic applications.

Awesome Hungarian NLP


Awesome Hungarian NLP is an up-to-date and ever growing collection of NLP resources available for Hungarian.

Open-source contributions

https://img.shields.io/github/stars/oroszgy?style=social https://img.shields.io/github/followers/oroszgy?style=social

Occasionally, I contribute to (mostly NLP related) open-source projects. I’ve made some improvements on spaCy, AllenNLP and textacy amongst others. For a complete list, please check my Github profile

PhD thesis

During my PhD studies, I’ve done extensive research in morphosyntactic processing of agglutinative languages and developed state-of-the-art NLP methods for Hungarian. [Read more »]


I’ve published several papers in scientific journals and conferences. For the complete list of publications check Google Scholar or MTMT.


As part of a research program with the University of Kent, I’ve enhanced Erlide with refactoring capabilities by integrating it with the Wrangler tool.