Building RAG Systems with Mixed Numeric and Text Data - Workshop

Explore the intricacies of building Retrieval-Augmented Generation (RAG) systems on datasets containing numerical data in this 59-minute workshop presented by Mór Kapronczay, Lead ML Engineer at Superlinked. Learn why search is challenging in real-world scenarios, understand the importance of RAG in leveraging Large Language Models for business applications, and discover how to effectively combine embeddings from different data modalities to create high-performing RAG systems. Follow along as Kapronczay demonstrates the process through an example of developing a chatbot for HR policies using Superlinked. The workshop covers an introduction to the topic, explains the difficulties of real-life search, delves into RAG's significance, showcases practical implementation using Superlinked, and concludes with a Q&A session. Gain valuable insights and hands-on experience in improving RAG performance through enhanced retrieval techniques.

- Introduction
- Why search is hard in real life ✅
- What is RAG and why it is important ✅
- Using Superlinked to build RAG ✅
- Q&A