Oracle AI Vector Search Benefits….

One of the biggest benefits of Oracle AI Vector Search is that semantic search on unstructured data can be combined with relational search on business data in one single system.

This is not only powerful but also significantly more effective because you don’t need to add a specialized vector database, eliminating the pain of data fragmentation between multiple systems.

These new capabilities also support Retrieval Augmented Generation (RAG) (https://www.oracle.com/artificial-intelligence/generative-ai/retrieval-augmented-generation-rag/), a breakthrough generative AI technique that combines large language models (LLMs) and private business data to deliver responses to natural language questions. RAG provides higher accuracy and avoids having to expose private data by including it in the LLM training data.

As an example, suppose you use an application that allows you to find a house that is similar to a picture you took of one you like that is located in your preferred area for a certain budget. Finding a good match in this case requires combining a semantic picture search with searches on business data. So you can create a table like this below,

CREATE TABLE house_for_sale (house_id NUMBER, price NUMBER, city VARCHAR2(400), house_photo BLOB, house_vector VECTOR);

and now, after inputting the data, you can query the table just like any other table in the database,

SELECT house_photo, city, price FROM house_for_sale

WHERE price <= :input_price AND city = :input_city

ORDER BY VECTOR_DISTANCE(house_vector, :input_vector);

In conjunction with Oracle Database 23ai, Oracle Exadata System Software release 24.1.0 introduces AI Smart Scan, a collection of Exadata-specific optimizations capable of improving the performance of various AI vector query operations by orders of magnitude.

AI Smart Scan automatically accelerates Oracle Database 23ai AI Vector Search with optimizations that deliver low-latency parallelized scans across massive volumes of vector data. AI Smart Scan processes vector data at memory speed, leveraging ultra-fast Exadata RDMA Memory (XRMEM) and Exadata Smart Flash Cache in the Exadata storage servers, and performs vector distance computations and top-K filtering at the data source, avoiding unnecessary network data transfer and database server processing.