Unleashing the Potential of Unstructured Data with LLMs and Databricks

Unstructured data can provide valuable insights for decision-making and problem-solving in various domains. However, extracting and analyzing such data can be challenging due to its complexity and diversity. This presentation will highlight the Retrieval augmented generation (RAG) architecture, an LLM framework the North Dakota University System (NDUS) has used for leveraging unstructured data, including policies and procedures, using the native AI features within Databricks. We will demonstrate how Databricks has been critical in creating an AI portal for all NDUS staff, faculty, and students to access their approved AI apps. We will also discuss the benefits and challenges of using the RAG architecture and Azure Databricks for unstructured data analytics and generation.