SAP HANA Vector Engine

The purpose of this lesson is to introduce one of the latest features of SAP HANA Cloud - the Vector Engine.

A vector datastore manages unstructured data - such as text, images, or audio - in high-dimensional vectors or embeddings, to provide long-term memory and better context to AI models.

SAP HANA Cloud will natively store and search so-called vector embeddings, which are numerical representations of objects, along with business data as part of its industry-leading multi-model processing capabilities to power intelligent data applications.

With these vector capabilities, SAP HANA Cloud will enable Retrieval Augmented Generation (RAG), facilitating the combination of LLMs with private business data. The SAP HANA Cloud vector engine will allow the storage, process, and analysis of vector embeddings in SAP HANA Cloud, allowing for a very efficient semantic search on data. These applications learn and adapt to new information, enabling automated decision-making and empowering developers to be the architects of the future.

Note: SAP HANA Cloud vector engine native SQL can also be used in a RAG scenario using the L2DISTANCE and COSINE_SIMILARITY functions. For more in-depth information and hands-on on this please register for a free SAP HANA Cloud Basic Trial.

Advantages of Multi-Model Architecture

• Integrate business data with graph, spatial, document, and vector data in a single platform
• Eliminate the need for numerous specialized database platforms
• Prevent data silos & data duplication efforts

The primary use case for a vector database is to enhance decision making processes by combining the results of a vector similarity search with other business data. A typical process flow for this scenario would look as follows:

Similarity Search Process Flow

The focus of this hands-on workshop is to demonstrate how to work with vectors in SAP HANA Cloud using SQL queries and functions. Similarity searches using pre-created vectors are used to determine suitable alternative products from a list of suppliers.

The actual process of creating vector embeddings and performing RAG searches using LLMs will be covered in a separate workshop.

The first step is to setup and configure the environment in order to complete the exercises. Please continue to the next lesson to get access to the HANA Database Explorer.