Summary
Erwin Marsi is a data scientist in Trondheim with 15 years of experience blending academic research in language and speech processing with industry-grade software development. Trained as a PhD in Language & Speech Technology, he has a deep background in NLP tasks such as named entity recognition, relation extraction, text generation and causal knowledge extraction from scientific literature. In industry he focuses on automatic detection of misuse of audiovisual streaming content, translating research-grade machine learning methods into deployable systems. His career bridges long-term academic projects across multiple European universities and practical product work at Sportradar, reflecting both theoretical depth and production engineering discipline. Notably, he couples expertise in prosody and speech synthesis from his PhD with modern ML pipelines, offering a rare combination of linguistic insight and applied data engineering.
15 years of coding experience
18 years of employment as a software developer
Doctor of Philosophy - PhD, Language & Speech Technology, Doctor of Philosophy - PhD, Language & Speech Technology at Radboud University
Dutch, English