View From the Top by Paul Moxon, Senior Vice President, Data Architecture and Chief Evangelist
A Logical Approach to Data Integration and Data Management
Artificial intelligence (AI) is top-of-mind today, and for good reason: It promises to revolutionize the IT industry in ways that will make the Internet seem quaint. To truly move the needle for organizations, AI applications need to be supported by a foundation of high-quality, governed data from multiple sources, delivered instantaneously in an actionable format. In order for a generative AI (GenAI) large language model (LLM) to intelligently answer a question about how many cars were sold by a company in a given week, the LLM needs immediate access to the company’s sales data, and in the right format.
Organizations can provide LLMs with the data they need via training, tuning, and prompt engineering, but these are gradual processes that accrue their benefits over time. Retrieval augmented generation (RAG) is a promising solution, as it provides a way for LLMs simply to reach out, when necessary, to additional databases. To support RAG, organizations need data that is always ready to be leveraged by an AI, and this is not always easy. This is where organizations lose heart in their AI journey, as the ability to provide data that is always “AI-ready” normally requires excessive amounts of time, effort, and expense.
Logical data management solutions, like the Denodo Platform, provide the quickest, easiest, and least costly way to support RAG-enabled GenAI applications. The Denodo Platform can be deployed as a universal data access layer above an organization’s data sources, providing AI applications with a steady stream of trusted data, in real time.
Denodo Technologies
www.denodo.com