The convergence of cloud native architectures, Kubernetes, AI, and modern data management is powering the next evolution of applications. Oracle Cloud Infrastructure (OCI) is here to expedite your journey.
Leo Leung, Vice President, OCI and Oracle Technology
The one thing about information technology that never changes is that it’s always changing.
What’s different going into 2023 is the accelerating rate of change. Granted, cloud computing came on fast, becoming the predominant IT deployment method over the last 16 years. But it’s taken just a few short years for Kubernetes, the open source cluster management system, to become the de facto operating system for new cloud native services.
Widespread adoption of Kubernetes, along with huge advances in artificial intelligence (AI), machine learning, the proliferation of data sources at our fingertips, and advancements in modern data management — have all come together to form the foundation for powerful next-gen “smart” applications.
Tomorrow’s apps are intelligent, delivering innovative customer experiences with the combination of cloud native services, real-time data, and democratization of AI.
We’re already familiar with predictive recommendations for what we should buy or click on next based on historic data. We’re also used to voice-enabled or computer vision-enabled applications- whenever we talk to our smart devices to check the weather or to play music, or when our car brakes after identifying a possible collision. Technological progress has opened the floodgates to advanced capabilities, enabling the creation of applications that once may have resembled something out of a sci-fi movie.
Think of the scale of infrastructure in the background that powers and directs all modern connected experiences and activities for billions of people, across millions of devices, in numerous interactions, interfaces and data points. What do tomorrow’s apps look like, and what does this mean for tomorrow’s cloud infrastructure?
Think of the growing number of data points, types of data, and the variety of sources and streams that need to be tapped, analyzed on the fly, enriched, cross-referenced, so that machines understand user queries, deliver the most optimized personalized results, as well as effectively employ AI algorithms — which are also, at their core, code that is applied to data sets, too.
Data and AI (as a use case of data), and the constant real-time “computation” between them and the user’s input – with one interaction feeding the next response – are the key value that next-gen apps provide. From personalized recommendations for your next watchlist, to ‘robots’ or AI investing our money for us, driving for us, or performing our next surgery.
The maturation and democratization of AI make these capabilities more accessible to a broader set of organizations, without requiring advanced AI scientist skills. For example, OCI offers pre-trained AI services with customizable ML models, that enable developers to simply connect to an end point to leverage speech/language recognition capabilities, computer vision, anomaly detection, AI forecasting and more—without requiring advanced skills.
Customers can also build their unique ML models and AI algorithms and train these more efficiently on high performance elastic cloud infrastructures. SoundHound is a big player here. Its AI technology runs real-time voice interactive systems used by Mercedes-Benz, Hyundai, Kia, Pandora, Netflix, and others. The Santa Clara, Calif.-based company uses an array of OCI services including high-performance GPUs, HPC capabilities and Kubernetes infrastructure to run more than 100 million queries per month.
Use of AI-fueled image searches is also skyrocketing. Snap Vision, for example, builds visual search tools used by retailers to help shoppers find products. During the pandemic, the London-based company gave U.K. retailers free use of its technology so shoppers could browse and buy clothing, even with physical stores shuttered. To meet the resulting surge in demand, Snap Vision switched its cloud service to OCI, successfully scaling to accommodate the influx of new business without interruption while cutting its costs 40% compared to its previous provider.
Oracle has spent years expanding and fine tuning what its databases can do. They now manage all types of converged data; come with embedded ML capabilities; and bring pre-trained AI models to be applied to that data—all features that come in handy in powering powerful, intelligent applications.
While many consumer applications tap into vast troves of publicly available data, remember that enterprises also are home to massive amounts of private business information and customer data as well. On average, Enterprises today have more than 400 data sources required for advanced analytics or other data-driven use cases. Many of the most mission-critical business applications still run on-premises, such as financial services, HCM, and ERP systems that continue to generate massive amounts of critical data that is often underutilized and should be more easily accessible for modern apps.
OCI and Oracle’s database expertise come in very handy for businesses wanting to make better use of this internal data. They can do so by implementing a REST API layer atop those applications to make pertinent data available for other approved uses. They can, for example, build AI chatbots to make it easier for people to perform common tasks based on this data, or leverage AI forecasting or other capabilities to Strengthen business efficiencies.
Oracle Autonomous Database, for example, as its name implies, automates management, security, and monitoring tasks allowing “hands free” database operations.
This database expertise is critical to helping ensure that data security, privacy, and usage are accomplished according to governance policies and regulation.
Distributed architectures like microservices, and the technologies used to run them – such as Kubernetes – are already more “chatty.” Modern experiences rely on almost constant communications between loosely coupled services to deliver the end-to-end customer experience.
Now, think about the additional network implications of the growth of data and AI-driven applications. Often, these use cases require increasingly faster – almost instantaneous – computations and low latency responses, with ‘weightier’ data-rich payloads having to continuously be transmitted between distributed services and the user.
For one thing, OCI’s ultra-low-latency, non-blocking resilient networking is particularly well suited to handling Kubernetes and microservices architectures, as well as data-heavy payloads across multicloud or distributed environments, while ensuring optimal performance.
OCI’s varied menu — which includes bare metal, high-performance computing (HPC), graphical processing units (GPU) compute options — brings the performance and scale to power these rapidly evolving applications.
Smart applications surround us. They run in our laptops, and in cyber-physical products like self-driving cars, robots, implanted medical devices, IoT, and more. That means constant, high-throughput, low latency communications between services now must happen across a growing and increasingly more distributed infrastructure – spanning cloud, multicloud, on-prem, edge, and embedded devices. This results in complex network topologies and bandwidth or resiliency constraints.
OCI deployment options allow customers to run a full range of cloud services in their own premises via OCI Cloud@Customer or in the OCI public cloud (or both). That deployment flexibility along with consistent pricing across all scenarios make OCI an ideal choice to run these modern applications.
With Oracle’s pedigree running some of the most demanding, sensitive, and large-scale apps in the world, OCI was architected with security safeguards built in by default, at no additional cost. These include tenant isolation, network security, threat intelligence, zero-trust role-based access control (RBAC) implemented throughout, and data security measures turned on by default.
Kubernetes, serverless functions and other technologies for running modern cloud native services will become even more intertwined with our inevitable data and AI driven future.
These apps would leverage advanced converged data stores, new modes of data operations, AI libraries and ML models, and more demanding requirements for fast computations and low latency responses – working together to support more smart use cases and innovative apps. OCI was designed to enable rapid development of apps today, while also enabling you to future-proof for tomorrow’s apps to be able to effectively leverage these strategic investments and new developments.
Advances across Kubernetes, AI, and data management are sparking the development of a raft of powerful new applications in both consumer and business realms. Those applications require modern infrastructure that offers optimal speed as well as the ability to securely handle reams of data in myriad forms.
OCI, both because of Oracle’s data management know-how and the fact that it is the most modern of the major public clouds, is the best platform for those thoroughly modern workloads.