A Vector Store is a system used to store data that has been converted into vectors using embedding techniques. This conversion allows unstructured data, such as free text or images, to be transformed into numerical representations (vectors) that reflect the characteristics of the original data. Once the data has been embedded and stored as vectors, it is possible to perform vector searches. This means that when an unstructured query is performed, it is also converted into a vector, and the system retrieves the most similar vectors stored, which facilitates the identification of information related to the query.
In the context of a RAG Assistant, a default Vector Store Provider is assigned according to the instance configuration. To configure it, go to the RAG Assistant.
The Globant Enterprise AI SAAS solution offers the following options:
Other available Vector Store Provider (available for on-premises installations)