Building Vector-Powered Products with User Embeddings

Explore the power of vector embeddings in building engaging, explainable, and safe software in this 42-minute conference talk from the Data Science Festival. Delve into the world of user modeling and its applications in recommender systems, bot detection, and more with Daniel Svonava, Co-Founder at Superlinked. Learn about the importance of centralizing user models in the Machine Learning stack, drawing parallels to the Customer Data Platform industry. Discover key concepts such as vector embeddings for user understanding, Vector Ops for production deployment, and the foundational role of user modeling in various applications. Gain insights from real-world experiences implementing the Superlinked framework in social platforms and marketplaces. Whether you're a generalist engineer looking to get started or a seasoned veteran seeking advanced tips, this talk offers valuable knowledge on leveraging vector-powered products to enhance user experiences and streamline data operations.