The infrastructure of AI – the technology that is enabling our imagination to become real

Hyperintelligence awaits

As progress in technology infrastructure marches rapidly on, more and more businesses are embracing artificial intelligence (AI) to enhance their competitiveness.

Improvements in connectivity, storage capacity and processing power at lower costs are enabling this revolution.

New developments in cloud, wireless networks, the Internet of Things, Application Programming Interfaces (APIs), quantum computing, and microservices – to name just a few – are working together to build the hyperintelligent enterprise.

In this chapter, we’ll look at the infrastructure approach needed for success.

The story so far

Digital technology has changed nearly every facet of our lives – but what’s enabling it?

Behind all of the solutions that recognize our faces, that tell us what and whom we are going to like, that provide instant search results and real-time electronic payments, lies the infrastructure of intelligence.

In the past, the mainframe was the mainstay of the infrastructure landscape. While that era should be long gone, many industries still rely on large, monolithic solutions running on infrastructures that are costly to maintain, rigid in their architecture, and too restrictive for the new digital age.

Hyperintelligent organizations are transitioning to an infrastructure that delivers the speed, responsiveness, and reliability modern consumers and employees expect – an infrastructure that can receive, store, and instantly recall the ever-increasing quantities of data that today’s AI solutions rely on.

The new landscape is characterized not by monoliths, but by nimble, modular technology. By adopting a “building block” architecture that supports continuous development and dynamic adjustment, resources can be added and removed quickly and affordably. This enables organizations to embrace intelligent solutions today, whilst future-proofing their infrastructure for the developments of tomorrow.\

The new age

Most businesses now use Software-as-a-Service (SaaS) solutions, hosted in the cloud. This has become the standard mode of software distribution, replacing on-premises installations using CD-ROMs and floppy disks.

However, there is a major difference between subscribing to SaaS applications and completely redesigning your infrastructure around what the cloud can offer.

Organizations that choose to go “native” with cloud can achieve a broader range of benefits and help transition to an AI-first future.  That’s because of the range of functionality that is now available on the cloud in the form of web services and microservices. 

Building block architecture

Constructing a complete business process or application can be as simple as selecting functional modules from a catalogue and assembling them in the same way you would put together Lego blocks.

These independent building blocks communicate with each other using APIs. Many APIs are freely available on the market, in an open and standardized format – enabling you to link to new applications easily and affordably. You can also use them to incorporate new microservices into an existing application chain, to get more value out of your existing enterprise solutions.

Microservices and APIs give confidence in the ability of constantly evolving solutions to continue to thrive in a constantly evolving ecosystem.


Continuous development

APIs eliminate the need to recode and retest a microservice every time it gets updated or replaced.
In fact, it can take as little as 15 seconds to initiate the production of a microservice or to execute an application update in a building block architecture.

Compare this to a traditional legacy environment, where it could take as much as 12 months of development and testing to update a core in-house business application.

Digital businesses update their critical applications millions of times each year using the modular flexibility of microservices and APIs.

 

Such dramatic shortening of production cycles allows for more frequent – even real-time – application development. This helps ensure key systems meet the evolving requirements of the business, the changing expectations of customers, and the latest technological advances.

Dynamic adjustment

A further benefit of the cloud and microservice-based architectures is the ability to procure dynamically physical resources based on exact workloads.

Today, most businesses still over-resource each application to ensure uptime. They clone their production environments to respond to a potential peak in demand. If that peak never happens, they are paying for unused resources. However, that has to be balanced against the risk that if demand exceeds the predicted maximum level, the core applications could fail – causing untold reputational damage and lost revenue.

Alternatively, microservice-based applications operate in their own independent software container, which determines the resources required for its own execution. When demand peaks, the required microservices are replicated automatically to take up the load. When the peak has passed, the duplicates are destroyed, and the resources are made available to other applications.

Today’s leading e-commerce and streaming sites use dynamic adjustment techniques for instantaneous load balancing to keep their services online even at peak times.

Don’t forget the data

When thinking about new forms of infrastructure, it’s important to remember that we are in a new age of technology, where data quality and security is everything.

Data becomes knowledge that provides valuable insight into business operations. Acting effectively on that insight can help an organization differentiate itself and create new value.

Data, knowledge, and insight are not static, but are constantly in flux. Any organization that is rethinking its infrastructure design must also consider how it will collect, process, exploit, and circulate them in a secure and effective way.

In the next article, we’ll explore how the use of AI allows organizations to shift their focus from efficiency to effectiveness, leading to far richer user experiences and better business outcomes.