How to align your LLM judge for better evaluations

In this webinar, we will cover how to build LLM applications correctly with a special focus on aligning LLM judge. While creating a RAG pipeline is now straightforward, aligning your LLM Judge for accurate evaluation remains a complex challenge. In this webinar, we’ll dive into various evaluation strategies with a special focus on aligning your LLM Judge. Using a RAG pipeline as a case study, we’ll demonstrate how to build an effective evaluation system with LlamaIndex and leverage W&B Weave for systematic assessment and annotation. Chapters: 0:00 Introduction and overview 1:36 Importance of evaluation in LLM Applications 3:33 Frameworks: LlamaIndex vs LangChain 4:14 Weights & Biases in LLM Ops 5:01 What is RAG? 7:01 Components of a RAG pipeline 16:44 Demo time 20:00 Building the retriever and query engine 27:01 Integrating Weave and viewing Traces 30:53 Customized evaluation and comparison 47:04 Final thoughts and summary Join us as we explore the entire evaluation lifecycle and uncover best practices.