Track A - Using Streamlit, Snowflake's Document AI and RAG to Enhance Customer Experiences

By utilizing RAG (Retrieval Augmented Generation), we can reduce hallucinations and "ground" a model response by making a set of relevant documents available to the LLM as the context to a question's response. This session will give a brief overview of how RAG works and then show how we can utilize the information contained outside of our database in unstructured documents (PDFs) to increase our confidence in our LLM's response. We will walk through the ingestion of these documents and the retrieval of relevant sections for use in a custom Streamlit application for use by our customer support team, financial advisors, or analysts.