Oracle AI Database 26AI: Revolutionizing AI-Driven Applications and Analytics

[Music] AI changes everything. Oracle AI database 26AI introduces hundreds of

major new features and thousands of enhancements. This is a monumental release that delivers breakthrough

capabilities in architectural simplicity across four key focus areas. Application development, analytic lakehouses,

missionritical workloads, and especially AI. With Oracle AI databases converged architecture, you can infuse AI into all

your mission critical applications, data types, and workloads to achieve breakthrough innovations and

productivity. One major focus area is making it dramatically simpler and faster for

developers to work with data at scale. Oracle AI database includes the ability to simply and efficiently query data as

graphs and the ability to store and access data as JSON documents with ultra high performance and MongoDB

compatibility. Further, Oracle AI database goes well beyond just providing the major data

types. It now unifies the relational JSON and graph data models. Now developers can get the benefits of all

the models at the same time. For example, an app can access data in the relational format using SQL while other

apps can access the exact same data as a JSON document or as a graph. This is a huge breakthrough unique to Oracle

that's going to change the way how developers access and process data to provide a full AI developer

experience. Oracle's next generation of development tools, including Apex AAI interactive reports. Apex AAI

application generator and SQL developer enable developers to rapidly generate schemas, SQL reports, and even entire

applications starting from natural language prompts. Others include JavaScript store

procedures, CFKA compatible APIs, property graph queries, and full consistency for stateless REST APIs.

These are just a few of the dozens of development focused features in Oracle AI database that will dramatically

improve developer productivity and innovation. Our second major focus area is AI

analytics on open data links. What if you could easily run sophisticated AI analytics across all the data in your

enterprise? Now with the Oracle Autonomous AI lakehouse and the Apache Iceberg open

table format, you can every major analytics platform is getting support for storing data lakeink data using the

industry standard Apache iceberg table format. Oracle Autonomous AI Lakehouse uniquely enables you to run the

industry's most comprehensive AI and analytic SQL on iceberg data with Exodata performance and pay-per-use

serverless scaling. You get everything great about iceberg data sharing plus everything great about Oracle Analytics

in a unified autonomous data lake while helping to avoid lock in data movement or security and compliance risk. These

are just a few of the dozens of features in Oracle Autonomous AI Lakehouse that will help you generate new business

insights and produce more value from all your data. Others include multi- vendor catalog federation, delta share and

spark integration. Another major focus area is making your mission critical data architecturally

simple and scalable across the entire stack. For example, with Oracle AI database 26AI, we've addressed caching

issues for developers. While there are many vendors out there that claim to offer caches deployed in the mid tier,

none of them are real caches. App developers still have to write code to manually populate and maintain the data

in the cache while doing the extra work of keeping it consistent with the back-end database. In contrast, Oracle's

true cache feature delivers a cache that's always consistent with the data in the database because changes are

automatically propagated to true cache in real time. Now, many of you are familiar with globally distributed

databases. They break down a single logical database into multiple physical databases either to enhance scalability,

availability or to support data sovereignty requirements by distributing data across different geographies.

With raft replication, Oracle AI database 26 AI makes it easy to deploy multimaster activeactive distributed

databases that can fail over in only about 3 seconds with zero data loss. So your most important data is protected no

matter where it's located. And to further safeguard your data, we've also addressed the growing risk posed by SQL

injection attacks. Oracle AI Database 26AI delivers SQL firewall built directly into the database itself to

help minimize latency and complexity. Unlike midtier SQL firewalls, it's been engineered to help prevent bad actors

from bypassing it. AI is reshaping every industry and redefining what's possible. That's why

Oracle AI database has AI at its very heart. But what you may not realize is that 20 years before AI became the

headline topic, Oracle customers were already training and running machine learning models right next to their data

with no extra servers or messy integrations. Historically, databases have been limited to querying structured

and semistructured data. Now with Oracle AI database, we've introduced AI vector search, a breakthrough technology that

enables searching for documents, images, videos, audio, and structured data based on their semantic content or meaning

rather than the words or pixels that make them up. And because AI vector search is built into the database, you

can easily run queries that combine both structured business data and unstructured data to provide accurate,

nuanced answers to realworld business questions. AI vector search makes it easy to implement capabilities like

personalized recommendations. You can recommend products by comparing a user's actions or purchase history to find

related products in your catalog. You can also easily combine AI vector search with LLMs to search for private

unstructured data that an LLM can use to answer natural language questions using a technique called retrieval augmented

generation. For example, you can answer customer support questions by searching product documentation and support

history to find answers that are semantically similar to the question being asked. While some AI solutions

become complex, costly and slow at scale, Oracle AI database enables missionritical AI at scale with Oracle's

unique engineered system, Exadata. By upgrading your Exodata software, you can offload AI vector queries to Exadata

intelligent storage, enabling them to run significantly faster. Vector offload works with traditional Exodata or the

new Exadata Exoscale software architecture which adds extreme elasticity and low cost to Exadata

making all Exadata benefits available for smaller workloads and businesses. Oracle has also made it easy for end

users to analyze business data using natural language. Oracle's natural language to SQL capabilities make it

fast and easy for developers and end users to generate complex SQL that answers analytical questions. With

Oracle's model context protocol or MCP support, LLMs can access your database to answer complex questions that require

the LLM to try multiple paths and ask for additional data as the reasoning process progresses. Schema annotations

explain the content of the tables to the LLMs, providing the context the LLM needs to understand your data. Using

these capabilities, you can now easily build AI agents that tap into live business data, run queries, apply AI and

ML models, and return results in real time. It's a much faster path from question to insight, and from insight to

action. These are just a few of the many AI features in Oracle AI database that will help you thrive in the age of AI.

Others include sophisticated vector indexes, select AI, full rag directly from SQL AI management and security

assistance. As you've seen, Oracle AI database 26AI is truly a monumental release with

hundreds of major industry-leading capabilities. We've engineered app development, analytics, and

missionritical to work together with AI so you can easily achieve breakthrough insights, innovation, and productivity.

Plus, you transparently benefit from every one of Oracle's missionritical capabilities. This means you can

immediately use AI for your enterprise apps of any size and criticality. Better yet, it's all available at no

extra charge on premises on all leading hyperscalers or in a cloud model in your data center. Thanks for joining us. I

hope you're as excited as we are about these amazing new technologies. [Music]