Roman Grebennikov is a seasoned data science engineer and startup enthusiast focused on building machine learning applications for the food industry. His work centers on search and recommendation systems, combining information retrieval, learning to rank, and large-scale data infrastructure.

Roman has shared his expertise with the data community through DataTalks.Club webinars, including Reinforcement Learning for Search, Building an Open-Source Feature Store with Apache Flink, and Practical Learning-to-Rank: Deep, Fast, Precise. These talks reflect his hands-on experience with ranking models, feature stores, and streaming pipelines in production settings.

He writes about machine learning engineering on his blog and shares his open source work on GitHub.