Summary
Vsevolod Svetlov is a Senior Staff ML Engineer with a decade of experience building deep learning systems for search, recommendation and video understanding, currently based in London. He specializes in metric learning and joint visual-text embedding spaces, applying sequence models—convolutional, recurrent and attention-based—to align video and short-text queries at scale. At Yandex he led teams that moved research into production, optimizing data pipelines, distributed and mixed-precision training, and delivering measurable online improvements. Equally comfortable in low-level optimized C++ and high-level TensorFlow research code, he combines production engineering rigor with research-driven model design. Outside work his interests span physics, e-commerce, choreography and stand-up comedy, reflecting a creative, interdisciplinary approach to problem solving.
10 years of coding experience
11 years of employment as a software developer
Specialist's degree, Machine Learning, Specialist's degree, Machine Learning at Московский Государственный Университет им. М.В. Ломоносова (МГУ) / Lomonosov Moscow State University (MSU)
English