Session Track: Data Intelligence
Session Time:
Session description
Navigating the deluge of daily news articles across diverse global regions presents a significant challenge for content platforms aiming to deliver relevant and engaging experiences. Traditional content discovery methods often fall short in providing the nuanced context and real-time connections users demand. In this lightning talk, the speaker will share how Glance, a leading content delivery platform, addressed this challenge by building a content knowledge graph powered by Neo4j. He will delve into the innovative architecture that ingests and processes over 50,000 articles daily, structuring them into a network of more than 50 million nodes and 150 million relationships. The session will specifically highlight how the system leverages the inherent power of graph relationships to identify and deliver real-time related article recommendations with sub-100 millisecond latency. You will discover practical strategies for designing and scaling a knowledge graph for dynamic, multigeographic content. You will learn how to harness native graph traversals and pattern matching to power sophisticated, real-time recommendation engines without relying on graph embeddings for current recommendations. Furthermore, you will gain insights into building high-performance, real-time graph applications that drive significant business impact and open doors for understanding all kinds of content and cross-content relations.
Machine Learning Engineer, Glance
Himanshu Aggarwal, a machine learning engineer, specializes in designing and developing personalized recommender systems and advanced AI solutions. With a robust background spanning research labs, media streaming applications, and content delivery platforms, he brings a strong blend of research, engineering, and applied data science expertise. Passionate about search and recommendation systems, Himanshu focuses on leveraging AI to optimize content understanding and generation. he is dedicated to defining next-generation user experiences across various domains.