Summary
Devavrat Tomar is a computer vision researcher based in Zurich with nine years of experience building ML systems across industry and academia, now focusing on anomaly detection in manufacturing at Cerrion. He completed a PhD at EPFL in data-efficient machine learning and medical image analysis after advanced studies at ETH Zürich and IIT Guwahati, blending strong theoretical grounding with production-minded engineering from multiple roles at Adobe. His research spans vision-language and foundation models, self-supervised learning, and domain adaptation, with a practical bent toward deploying robust, data-efficient solutions in real-world settings. Notably, his trajectory bridges long-term research during a doctoral program with earlier hands-on software development and ML internships, giving him rare fluency in both prototyping and scaling vision systems.
8 years of coding experience
3 years of employment as a software developer
Bachelor of Technology (B.Tech.), Electronics and ELectrical Engineering (major) Computer Science (minor), 9.06/10, Bachelor of Technology (B.Tech.), Electronics and ELectrical Engineering (major) Computer Science (minor), 9.06/10 at Indian Institute of Technology, Guwahati
Master's degree, Data Science, Master's degree, Data Science at ETH Zürich
Doctor of Philosophy - PhD, Data Efficient Machine Learning, Computer Vision, Medical Image Analysis, Doctor of Philosophy - PhD, Data Efficient Machine Learning, Computer Vision, Medical Image Analysis at EPFL (École polytechnique fédérale de Lausanne)