The Ultimate Guide to Oracle's Cloud & AI Database
The advances include vector search, natural language queries, low-code AI development, and colocation with Microsoft, Google Cloud, and AWS.
Welcome to the Cloud Database Report. I’m John Foley, a long-time tech journalist, including 18 years at InformationWeek, who then worked in strategic comms at Oracle, IBM, and MongoDB. I invite you to subscribe, share, comment, and connect with me on LinkedIn.
It’s been the year of AI in the database market, with LLMs, GenAI, vector embeddings, RAG, copilots, and agents showing up everywhere.
No one has been busier than Oracle. Since January, Oracle has introduced LLMs/GenAI for Oracle Cloud Infrastructure (OCI) and its Fusion apps; AI vector search; GenAI-powered analytics, GenAI agents for RAG development, and a “zettascale” AI supercomputer. (Zettabytes are 1,000X exabytes.)
It has also released two AI-packed databases: Oracle Database 23ai and HeatWave GenAI, which features LLMs right inside the database. And perhaps most importantly, Oracle struck colocation agreements with Microsoft, Google Cloud, and AWS (in that order) — knocking down the Berlin Wall of cloud databases.
At the same time, Oracle announced partnerships with NVIDIA, OpenAI, and other major players. And it’s building out data centers with tens of thousands of GPUs, addressing one of the biggest issues gating the pace of AI development — training and processing capacity.
Throughout the year, I’ve had occasion to talk directly to the Oracle experts who are leading much of this product development. My analysis appears regularly on the Oracle Connect website, where the company shares customer stores and articles on its technical advances.
With Oracle AI, the whole is greater than the sum of its parts. So here I’ve pulled together some of my articles to provide a more complete picture.
1. Oracle Database 23ai
Trustworthy AI requires good, clean, accurate data. Oracle has been refining the art of data management for 45 years, so it only makes sense that Oracle Database would provide a strong foundation for that. And now all the more so since Oracle has beefed up its AI capabilities with the mid-year release of Oracle Database 23ai.
Here’s what I wrote:
“AI initiatives need a true AI database. And while the market is crowded with DBMS platforms, few offer all the pieces needed for AI workloads: low-code development, built-in machine learning, natural language queries, optimized storage, and high performance caching….Oracle has answered this challenge with Oracle Database 23ai, which incorporates thousands of enhancements and more than 300 new features.”
See the full article: “Oracle Database 23ai sets new industry standard for a modern AI database”
2. Multi-cloud data & AI
Oracle’s data center colocation agreements with the big three hyperscalers —Microsoft, Google Cloud, and AWS — open new possibilities for how businesses manage data in and across clouds. The benefits include choice of cloud service providers, flexibility in workload distribution and management, and split-second performance.
Here’s what I wrote:
“Not only do customers gain access to Oracle AI innovations in AWS, Azure, and Google Cloud, but it’s a two-way street….Developers can tap into Oracle Databases using Google Cloud services, such as Gemini foundation models. In Azure, they can mingle Azure AI services, like OpenAI large language models, with Oracle enterprise data. Or in AWS, they can seamlessly connect enterprise data in their Oracle Database to AWS’s AI and ML services, including Amazon Bedrock.”
See the full article: “New cloud partnerships put your data where you need it”
3. Vector search
AI development teams must choose between two ways to store, administer, and retrieve vector data: a purpose-built vector database like Pinecone or a multi-modal database such as Oracle’s. The term Oracle uses is a “converged” database. Here’s why: Oracle 23ai supports vector data, plus blockchain, graph, spatial, JSON, REST, events, IoT streaming, and more—all as part of the core system.
Here’s what I wrote:
“As vector-enabled applications advance from pilot projects to customer-facing deployments, they must deliver the levels of performance, scalability, security, and reliability that business managers expect from run-the-business applications. Oracle AI Vector Search clears that bar by leveraging other enterprise-class Oracle capabilities, such as Real Application Clusters (RAC), partitioning, sharding, security, analytics, and disaster recovery.”
See the full article: “5 advantages of using an integrated vector database for AI development”
4. Natural language queries
An innovative capability called Select AI offers a powerful way to query Oracle databases using natural language. It does this in conjunction with popular large language models (LLMs), including Cohere and Llama-2.
Here’s what I wrote:
“Select AI uses LLMs to translate a natural language question into a SQL query that the database understands. The breakthrough is that, because LLMs have been trained to infer user intent, they can often interpret a casual or colloquial query and return the answer or content you’re looking for. Select AI, introduced in September 2023, brings this intuitive user experience to Oracle Autonomous Database. In fact, it’s able to scour your entire data estate, including other data sources, object stores, and data lakes, to find the answer to a query.”
See the full article: “Natural language queries to Oracle Autonomous Database? Yes—with Select AI”
5. Low-code development
Oracle’s APEX is one of the tech industry’s original low-code/no-code development environments. It’s been around for more than 20 years, is used by 850,000 developers, and now comes with new AI-enabled capabilities. Oracle says APEX has been used to build millions of apps.
Here’s what I wrote:
“The new APEX 24.1 release, introduced in June 2024, features dozens of enhancements, including three that stand out: APEX AI Assistant, which enables natural language queries; a “create application” assistant that generates blueprints with the attributes you want; and conversational AI dialogs for creating natural language user experiences.”
See the full article: “The power of Oracle Apex and GenAI for businesses”
6. HeatWave GenAI
Oracle is synonymous with its flagship Oracle Database, but it has other databases, the best known of which is HeatWave, which has been transformed from its MySQL roots into a state-of-the-art multi-cloud DMBS with integrated AI capabilities.
MySQL landed in Oracle’s lap in 2010 with its acquisition of Sun Microsystems. Oracle continues to put development resources into HeatWave due to the sizable MySQL installed base. HeatWave is an easier alternative for existing MySQL users who may want/need an enterprise-class database, but don’t want to bite off a full-fledged migration to Oracle Database.
Here’s what I wrote:
“The net result of having AI-enabling technologies baked into HeatWave GenAI is the ability to skip step-by-step development. Say you want to create a vector store for similarity search. That could require as many as nine distinct actions: discover documents, parse data, extract metadata, segment data, choose embedding model, create vector embeddings, and so on. HeatWave GenAI does it all, end-to-end, as an automated process.”
See the full article: “HeatWave GenAI eases development with integrated LLMs, vectors, and more”
