Enterprise Vector Store is designed to store, index, and search high-dimensional vector embeddings efficiently. Vector embeddings are numerical representations of data (text, images, audio, and so on) in a multi-dimensional space. Each embedding is a vector (list of numbers) that captures relationships between data points based on semantics or content similarity.
Vectors are an integral foundational component of large language models (LLM). They are an additional data type that can use Teradata’s parallel architecture to outperform other vector stores.
For detailed information about Teradata Vector Store, see Teradata® Enterprise Vector Store User Guide.
For directions to use the Lake console to work with vector store, see the following: