Enrico Zimuel

In this presentation, we will introduce the Retrieval-Augmented Generation (RAG) architecture. This technique uses natural language capabilities of Large Language Model (LLM) to query a private database. You can literally ask question about your data, without the need to fine-tune an LLM. In the lab session we will design a RAG system in Python using Langchain, an open source framework. We will store private documents in a vector database (i.e. Elasticsearch) and ask questions using natural language thanks to the usage of an LLM (e.g. ChatGPT).