Dávid Márk Nemeskey

Dávid Márk Nemeskey has been working with NLP for 15 years in both academic and industrial settings. He wrote his doctoral dissertation as a researcher at MTA SZTAKI. The huBERT model he published is perhaps the most popular Hungarian-language encoder for text processing tasks. From 2020 to 2026, he was a member of the National Laboratory for Digital Heritage at ELTE, serving as its Technical Director from 2024. He is also a member of the organizing committee of the Hungarian Conference on Computational Linguistics.

In 2025, together with researchers from the Laboratory and ELTE’s Department of Artificial Intelligence, he released Racka 4B, the first Hungarian reasoning LLM.