As we continue exploring the exciting developments shared at Oracle CloudWorld 2024, let’s take a closer look at the innovations in database technologies, AI, and security that are reshaping the future of cloud computing.
Andy Mendelsohn, Executive Vice President of Database Server Technologies, talked about 23ai, Autonomous Database, GenSQL, and APEX. Andy first explained the Oracle Database Vision to build Modern Apps, Analytics, and AI to use at every scale needed. The first key is the Converged Database for any database workload development, whether it's JSON, a need for a Vector Database if you need to do something that includes mapping, or a need for AI using many major algorithms; all of it will work on the Oracle converged database. Something worth noting is that Larry Ellison, just a few months ago, introduced the Oracle Vector Database and said it would be absolutely FREE with the Oracle Database. No extra charges for JSON, Objects, Spatial, Vector, Key-Value...etc. Andy showed that it's easy to add a vector data type to tables (vectorizing a blob image - binary large object) so that you can now search for the image vector with SQL or JSON. The second key part of how they're going to deliver on that vision is the Autonomous Database.
The Autonomous Database has the full Converged Database in it, but it runs fully managed, automatically tuned, backing itself up, recovering itself, and puts a patch on before you even know you need a patch. The Converged Database lets developers have the innovation they need to use any tool out there. Oracle allows the converged database to be accessed from APEX, Rest Services, Mongo, SQL, JSON, etc. You can also use their data catalog to transform or load data, use notebooks for Machine Learning, Auto Machine Learning (AutoML), Graph Studio, and more, all included in the converged database. Andy showed that instead of using a single-use proprietary database such as Aurora, DocumentDB, Redshift, Neptune, etc., you can now use that one converged Oracle database from AWS. Andy shared that Oracle's own software as a service runs an autonomous database with 79 different cloud services, all on autonomous today. Customers using autonomous databases include Siemens, Vodafone, Experian University, Accenture, NEC, Xerox, and Wilson. Andy then brought in a senior New York Stock Exchange (NYSE) director to explain how he uses Oracle. NYSE does $500 billion to a trillion dollars in trades per day on Oracle while running Exadata, RAC, Data Guard, Security, and Partitioning with a hundred terabyte database. In the future, NYSE wants to add Oracle's JSON tools for their developers, True Cache (consistent blocks at the database and app server level) for speed, and the Oracle Vector Search.
Then, Andy talked about SelectAI in the Autonomous Database. SelectAI generates SQL from a natural language using your preferred language (while writing queries to your database, which you can validate) without needing to understand where your data is or how your data is stored to gain insights.
Andy then brought out another customer, Vodafone, with Martin Block, head of Oracle cloud services at Vodafone, and Armon Ashura, Head of OCI (Oracle Cloud Infrastructure) modernization. Vodafone has 330 million customers in the European Union and Africa, with 10,000 plus databases. There are six dedicated public Cloud regions that are only used by Vodafone. They were migrating databases to Autonomous Database (where possible), and they were able to move 30% of current databases to Autonomous. With new databases, they decided autonomous-first (if possible), and of new databases, 100% ended up on autonomous databases. They used the autonomous database, a fully managed cloud service built for extreme performance, availability, scalability, and security. They use the fully managed converged Oracle Database with all the best practices on a management platform to give them the lowest total cost experience.
Andy showed a demo of the new multi-cloud that Larry talked about on Azure, Google, and AWS (in December). Maybe the most impressive thing I saw was when Andy brought Tim Klein from Oracle on stage. He ran a demo where he went into Microsoft Azure. He then created an Oracle autonomous database using the same prompts that people are using in Oracle: giving it a name, putting it in a region, either autonomous data warehouse or autonomous transaction processing, the version of the database, storage size, the number of CPUs, whether you wanted scaling or not, a username and password. He created that autonomous database, and you saw it in an Azure look, but you could then click on "go to OCI" and see it just as you see an autonomous database in Oracle. If you have never used Azure but are familiar with building an Autonomous Database in OCI, you could build it and then click on an OCI button with all of the database actions you do with Oracle Cloud now. You could look at SQL, Data Modeler, REST, JSON, Charts, APEX, and Graph Studio (just like on Oracle, but all running on Azure). I also could use SelectAI with that autonomous database. They showed a demo using SelectAI asking, "what are our total streams break out by genre and customer segment," and it returned the result of that query. I could say, "give me the top five customer segments and genre combinations by total views and rank the results," and it'll find that result for you; you can also ask it to show the SQL it used if you want. They are bringing together the best from Oracle and Microsoft to give you the Oracle autonomous databases with the highest availability, fully automated, fastest performance, and greatest agility, plus the things you might like on Azure, like virtual machines, OpenAI, and your DevOps with Kubernetes.
Andy also discussed some of the new Oracle Database 23ai features for developers, analysts, and DBAs. He showed how bringing AI to your data with algorithms, AI Vector search, SelectAI, and Retrieval Augmented Generation (RAG) can be used. Developers can use JSON Relational Duality to either look at a table using JSON or view it as relational with SQL. Oracle only stores it once and creates a view for JSON. 23ai also has critical features like True Cache, where I have a multi-tier cache, but the cache in each and every place, whether it's in the database, in-memory, or on the app server, is kept up to date (consistent). This uses a process very similar to how the buffer cache works. They also added a database firewall to prevent SQL injections. The 23ai Vector Search allows you to search for an image, a video, or a document. You do this by just adding a vector data type as a table column. In 23ai, they also added how operational property graphs could be used to see connections between connected data (such as bank transactions flow to find potential fraud).
Other Innovations in 23ai include the Oracle Vector Search, JavaScript stored procedures, and JSON relational duality.
Andy also talked about the new AppDev product (more in the next part of the blog post, in Juan Loaiza's talk). Oracle built it to accelerate the application development environment. Oracle's vision is to say what you want (in human language), and AI generates that application. In summary, Andy explained that Oracle's first goal is to allow you to put everything on Oracle's autonomous database, in the cloud, multi-cloud, or on-premise. Second, to use Oracle Database 23ai as the long-term release leveraging the converged database.
The announcements and discussions at Oracle CloudWorld 2024 solidified Oracle's position as a powerhouse in the cloud and database landscape. With cutting-edge developments in AI, security, and Autonomous Database, Oracle continues to drive innovation that empowers organizations at scale. As we move into the next section, we'll dive deeper into the next wave of advancements, including Oracle's new application development tools and how they're reshaping how developers approach building scalable, secure applications. Stay tuned for more insights into Oracle's transformative offerings.