Oracle HeatWave GenAI is now generally available. The release features the industry’s first in-database large language models (LLMs), automated vector store, scale-out vector processing, and natural language conversations informed by unstructured content.
The release enables enterprises to leverage generative AI with their data without needing AI expertise or moving data to a separate vector database. Available at no extra cost to HeatWave customers, these capabilities enhance data security and performance while reducing costs.
Key features include:
- In-Database LLMs: Simplify the development of generative AI applications, allowing for data search, content generation, and retrieval-augmented generation (RAG) with HeatWave Vector Store.
- Automated Vector Store: This enables easy use of generative AI with business documents without data transfer and AI expertise, automating the creation and embedding process.
- Scale-Out Vector Processing: This technology provides fast and accurate semantic search results using a new VECTOR data type and optimized distance function, enabling efficient parallel processing.
- HeatWave Chat: A Visual Code plug-in for MySQL Shell, offering a graphical interface for natural language or SQL queries and maintaining context for continuous, accurate conversation.
This Research Brief takes a deeper look at what Oracle announced.
