RAG and RecSys - Parallels in Retrieval Quality Improvement

Explore the parallels between Retrieval Augmented Generation (RAG) and recommender systems in this lightning talk from the MLOps.community AI in Production series. Discover how RAG pipelines are evolving to incorporate features similar to recommender pipelines, such as hybrid search and reranking. Learn how to implement hybrid reranking with LanceDB to improve retrieval quality. Gain insights from Chang She, CEO and Co-founder of LanceDB, who brings nearly two decades of experience in building tools for data science and machine learning, including co-authoring Pandas and developing personalized recommendations at TubiTV.

Intro
Welcome
About me
RagSys
Production
Retrieval Quality
Content Recommender
Feedback Mechanism
Limitations
Outro