Build a Retrieval Augmented Generation (RAG) based LLM assistant using Streamlit and Snowflake Cortex