Summary
Marija Stepanović is an Associate Data Scientist and applied AI researcher with eight years of experience developing fair, humane, and inclusive language and speech technologies informed by linguistics, cognitive science, and neuroscience. Her PhD work focused on ASR for low- and zero-resource languages, building multilingual text/audio datasets and fine-tuning large speech and language models for ASR, TTS, translation, and NLU tasks with parameter-efficient techniques like LoRA/qLoRA. Based in Copenhagen, she brings hands-on expertise in EEG, speech, and text signal processing, scalable multi-GPU training on HPC clusters, rigorous error analysis, and ethical bias mitigation. She also has a strong teaching and supervision record and a background in translation and language education, which gives her a rare practical perspective on language diversity and usability in AI systems.
7 years of coding experience
14 years of employment as a software developer
MA, English Language and Literature, MA, English Language and Literature at University of Niš
PhD, Computer Science, PhD, Computer Science at IT-Universitetet i København
MSc, IT and Cognition, MSc, IT and Cognition at Københavns Universitet