The RAG systems, such as Microsoft’s Copilot and Google’s Bard, are playing an increasingly crucial role in our digital daily lives, used every day by hundreds of millions of users.
As these giants compete for supremacy in the field of artificial intelligence, is there an emerging path towards an open-source RAG system that respects data privacy?
This presentation aims to answer this question by exploring the use of the Langchain library and OpenSearch as vector databases. We will begin with a quick analysis of the foundations of the RAG architecture. Then, we will focus on the implementation, also examining other alternatives to Langchain and OpenSearch. Finally, we will conclude by presenting the operation of a fully open-source and local RAG system within a concrete use case.
