Measuring and Improving Retrieval in RAG Systems - Haystack US 2024

Explore a conference talk from Haystack US 2024 focusing on enhancing the retrieval aspect of Retrieval-Augmented Generation (RAG) systems. Delve into Scott Stults' presentation on measuring and improving the quality of text chunks used for context in RAG. Learn about the challenges of traditional human expert judgments and discover an innovative approach using Large Language Models (LLMs) to generate quality assessments. Understand how this method can be applied to tune retrieval processes and ultimately enhance the quality of generated responses. Gain insights from Stults' extensive experience in implementing search solutions with complex security requirements across various defense and commercial projects. Benefit from the expertise of OpenSource Connections, a company specializing in Solr since 2007, and their best practices for securing Solr implementations.

Haystack US 2024 - Scott Stults: Measuring and Improving the R in RAG