Large Language Models (LLMs) are advanced artificial intelligence (AI) models that possess the remarkable ability to comprehend and generate human-like text. This capability empowers numerous natural language processing applications. However, a constraint of LLMs is that they are limited to the information they were trained on and cannot access new data after their creation. Retrieval Augmented Generation (RAG) overcomes this limitation by enabling LLMs to fetch relevant facts from a knowledge base and enhancing their functionality.

This solution is designed to provide users with simple answers to complex questions grounded in the organizations current knowledge base. It leverages the data science pipelines in OpenShift AI to ingest the data periodically. It also offers advanced features such as choice of different LLMs, different vector databases or catalogs to choose, along with options to change the hyperparameters within the user interface.