Taja Kuzman Pungeršek has been employed as a research assistant at the Department of Knowledge Technologies (E8) at the Jožef Stefan Institute since 2021. In the same year, she also began her doctoral studies in the Information and Communication Technologies program at the Jožef Stefan International Postgraduate School. Her research primarily focuses on natural language processing and machine learning.

She participates in numerous national and international projects. Notably, within the MaCoCu project (2021–2023), she contributed to building large collections of web texts for several European languages and ensuring their quality. In the ParlaMint (2021–2023) and ParlaCAP (2025–2027) projects, she was involved in compiling parliamentary text corpora that include speeches from the parliaments of 29 European countries and regions. In ParlaMint, she led the machine translation of over eight million speeches into English, and in ParlaCAP, she developed a topic detection model for parliamentary speeches. She is also involved in the Slovenian project MEZZANINE (2022–2025), focused on speech research, speech technologies, and speech language resources for Slovenian, as well as in the EMMA project (2023–2026), aimed at developing language technologies for the media industry, within which she developed a model for automatic topic classification of news content. Recently, she has become engaged in projects aimed at the development and evaluation of large language models, such as LLM4DH (2024–2027), LLMs4EU (2025–2028), and the LLMs4SSH knowledge centre of the CLARIN ERIC infrastructure (2025–), where she is working on evaluating large language models’ capabilities on various tasks in Slovenian and other South Slavic languages and dialects.

In addition to project work, Taja Kuzman Pungeršek is highly active in the Slovenian language resources and technology infrastructure CLARIN.SI, where she serves as a member of the executive board. Together with Nikola Ljubešić, she co-leads the CLASSLA knowledge centre for South Slavic languages, which offers educational materials on accessing language resources and technologies for South Slavic languages and also develops its own resources and technologies. Within this centre, Taja also helps organize workshops on the use of large text collections for linguistic research and maintains infrastructure for the iterative collection of large web-based text corpora, used for large language model development and linguistic research.

Her work has been published in several international journals and conferences in the fields of computational linguistics and natural language processing, including LREC, EACL, and COLING. In the five years since joining the Institute, she has authored or co-authored five scientific papers, twenty conference contributions, and more than seventy language resources and technologies. In 2024, she received the Best Paper Award at the Slovenian Conference on Artificial Intelligence. Her contributions have been cited over 570 times, and the models she developed - especially those for news topic classification and automatic genre identification - have been downloaded more than 650,000 times in total.

As part of her doctoral studies, she is researching automatic genre identification in web texts under the mentorship of Nikola Ljubešić. She is currently in her final year of study. During her studies, she has published three scientific articles indexed in Web of Science and Scopus. She has passed all exams with an average grade of 10.0 and is currently completing her doctoral dissertation, which she plans to defend by the end of the current academic year.

Taja is also an active member of the Department of Knowledge Technologies, where she has served for several years as one of the coordinators of visits from primary and secondary schools and has presented the work of the department during the Jožef Stefan Institute Open Day. Additionally, as a doctoral student, Taja served as a student representative of the Information and Communication Technologies program in the Student Council of the Jožef Stefan International Postgraduate School, where she participated in school governance and organized extracurricular activities to foster student engagement and social interaction. For her work, she received a special award from the International Postgraduate School in 2025.