Explore the latest approaches to building recommender systems using Large Language Models (LLMs) in this 12-minute talk by Sumit Kumar, an MLE at Meta specializing in Recommender Systems. Discover how researchers are transforming various recommendation tasks into language understanding or generation templates. Gain insights into recent developments in this field and learn how LLMs are revolutionizing the way recommender systems are constructed. Benefit from Kumar's extensive experience in RecSys at TikTok, NLU at Amazon, and speech processing at Samsung as he shares valuable knowledge on integrating LLMs into production-ready recommender systems.