Enterprise Vector-based Information Retrieval & Prompt Engineering at Scale using Open Source on AWS

Foundation Models perform well at natural language tasks such summarization, text generation and question answering on a broad variety of topics but struggle to provide accurate answers and without hallucinations about content that they haven't seen in their training data. Retrieval Augmented Generation (RAG) is a popular approach to mitigate this problem. When building next generation Information retrieval (IR) applications using a RAG architecture, enterprises encounter challenges of scale and operational complexity. In this talk, we will address these challenges and walkthrough a set open source technologies on AWS for building enterprise vector-based IR applications.