Oracle has launched HeatWave GenAI, a new offering that includes various advanced features such as large in-database language models, automated vector warehousing, scalable vector processing, and contextually informed natural language conversations for unstructured content. This allows companies to utilize generative AI capabilities for their data without the need for AI expertise or transferring data to a separate vector database. HeatWave GenAI is now available in all Oracle Cloud regions and other clouds.
Developers can create a vector store for unstructured enterprise content with a single SQL command using the built-in embedding models, thereby simplifying natural language searches using LLM in the database. This eliminates the need to provision GPUs and ensures the data remains within the database, which helps in reducing application complexity, improving performance, enhancing data security, and cutting costs.
New features of HeatWave GenAI include LLM in Database, which streamlines the development of generative AI applications by allowing customers to generate or summarize content, search data, and perform augmented recovery generation using HeatWave Vector Store. Automated storage of vectors in the database through the Vector Warehouse enables businesses to utilize generative AI with their documents without moving data to a separate vector database. Expanded Vector Processing supports semantic queries using standard SQL and can be parallelized across up to 512 HeatWave nodes for faster query responses.
Talk about heat waves is a Visual Code plugin for MySQL Shell that provides a graphical interface to HeatWave GenAI, allowing developers to ask questions in natural language or SQL. It features a built-in Lakehouse Navigator for selecting files from object storage and creating a vector store, while maintaining context with the history of questions asked. This facilitates contextual conversations and enables users to verify responses generated by the LLM.
Edward Screven, chief corporate architect at Oracle, emphasized that the integrated and automated AI enhancements in HeatWave GenAI enable developers to build rich generative AI applications quickly, without requiring AI expertise or data movement. This intuitive solution allows users to interact with their business data efficiently and obtain precise answers tailored to their business needs.
Article Source
https://www.ndtvprofit.com/amp/technology/oracle-announces-heatwave-genai-featuring-in-database-llms-automated-vector-store-2