This infographic explains how generative AI uses different kinds of data, from unstructured media like text, images, audio, and video to structured organizational data. Unstructured data is vectorized into multidimensional numerical representations stored in vector databases, enabling similarity search and model training. Semi-structured data, such as text within tables, can be extracted or concatenated for vectorization, enriching chatbots and knowledge tools. Structured data like metrics and transactions can power natural language BI through text-to-SQL, letting users query data conversationally. Together, these approaches make generative AI more effective for business insights and automation.

Join for free to read