Infrastructure modernization is key to AI success

As the world’s third largest artificial intelligence market (£72.3bn in 2024), the United Kingdom is home to 37,000 AI companies that employ more than 60,000 people and contribute about £3.7 billion to the economy.

The U.K. Government is also throwing its weight behind the technology with several initiatives and investment programs, including the AI Opportunities Action Plan to drive the use of AI tools in public offices.

Unfortunately, challenges such as legacy systems and data quality issues in public sector organizations are thwarting these plans. Indeed, legacy infrastructure is a major barrier to AI adoption in several U.K. industries, from the traditional construction sector to the rapidly digitizing healthcare industry.

Tech modernization imperative to run AI solutions

The latest AI solutions process hundreds of petabytes of data in wide-ranging formats, including text,  image, audio, video, and gesture in real-time; machine learning models analyze enormous datasets to uncover patterns and draw insights; predictive analytics solutions forecast events to enable businesses to respond proactively to external triggers; agentic AI tools learn from ongoing customer interactions, recommending next-best actions and taking decisions autonomously while the conversation is still on.

Highly scalable, intelligent and secure technology infrastructure is absolutely essential for supporting the massive compute, storage, and data processing capabilities required by these applications, on demand. Packed with these virtues, cloud is the preeminent infrastructure modernization solution.

This is why progressive enterprises are modernizing their core systems with cloud to transform operations across the board. For example, manufacturers are automating production activities across assembly, quality control, maintenance and other operational areas.

They are able to integrate AI within their supply chains to gain real-time visibility and thereby improve inventory management, identify bottlenecks, monitor vendor compliance, track sustainability metrics, and so on.

Examples include a luxury car and engine manufacturer, who have adopted a popular cloud platform to run generative AI and machine learning solutions to improve productivity and cost efficiency in engine design, turbine production, and other operations.

Energy companies have migrated siloed systems to a cloud-native platform to create a digital foundation for the company’s energy transition, which enabled them to deliver smart metering and customer solutions such as solar power, battery, EV charging etc. 

Modernized infrastructure takes AI from experiment to enterprise

A leading research and market intelligence firm found that on average, only 48 percent of AI projects progress to the production stage; they also said that it takes eight months for a prototype to get into production. It is widely reported that 50 to 70 percent of AI projects get stuck in the dreaded “pilot purgatory”.

One reason for this is the difficulty of integrating AI with existing systems and processes.

Although a 2024 survey suggests generative AI projects might be faring better, with 47 percent of 2,500+ firms with revenues exceeding  $10m making it from concept to rollout within 6 months, 38 percent of U.K. respondents admitted that they were struggling to scale gen AI from pilot to production.

A major hindrance was poor data quality and management – without accurate, relevant and adequate data, generative AI models cannot perform.

This is yet another reason why U.K. businesses should modernize their infrastructure.

A modern digital core running on cloud breaks data silos and supports AI integration – as well as training of gen AI models – by offering a centralized, accessible platform for storing, managing and sharing data in real-time.

What’s more, this improves collaboration between different teams and systems to facilitate ideation and innovation. Cloud is also the perfect environment for Agile software development; it offers all the required infrastructure, storage, networking and services resources on demand to support iterative development. 

Enterprises experimenting with AI models can easily test different scenarios in real-time on cloud, while running development and rectification activities in parallel.

Agile software developers working on new AI solutions can collaborate with co-workers to create rapid prototypes and take them to customers for their feedback to accelerate innovation.

Finally, the cloud platform closes the loop by providing virtually unlimited infrastructure and computing capacity to scale the AI pilots across the enterprise in quick time.

Cloud platforms protect AI and its data

AI can unlock unlimited possibilities for U.K. enterprises, but it also throws the threat landscape wide open. Legacy systems are simply not geared to protect the vast data used by AI technologies from increasingly sophisticated forms of attack.

Modernization addresses this challenge by replacing outdated, inadequate security tools with the latest security solutions. Cloud service providers have invested in robust cybersecurity features, such as encryption, access control and threat detection to protect workloads.

Offering centralized security management, continuous monitoring and automated responses, hyperscalers help enterprises reduce the risk of breach through early threat detection.

Last but not least, cloud’s in-built redundancy and disaster recovery capabilities ensure that businesses can continue to run their AI systems even when their physical premises are hit by calamity or contingency.

For highly regulated industries, such as healthcare and financial services, which are subject to data sovereignty rules, the ideal infrastructure for running AI is private cloud. In September 2024, Blackstone announced a £10 billion investment in an AI data center in Blyth, Northumberland.

Companies without their own data centers/ private cloud infrastructure can avail the services of companies that are investing billions. As many are investing billions in building sovereign data centers in the United Kingdom.

We've featured the best AI website builder.

This article was produced as part of TechRadarPro's Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro