Unlocking future AI-driven innovation by modernising yesterday’s legacy systems

Amidst evolving technological disruptions, legacy systems risk slowing down the pace of innovation for businesses. As these are often the technical backbone of their operations, it’s critical to update these systems if companies are to remain competitive, adapt to changing market trends, and drive operational efficiency.  

“The big problem with legacy modernisation has always been that it’s very hard just to understand what a system does today, particularly if it’s old or even inaccessible,” says Mike Turner, global leader of the Software and Platform Engineering service line at Cognizant. “Documentation will be out-of-date, and we could be looking at millions or even tens of millions of lines of legacy code.” 

The risks generated from sluggish legacy systems are significant. Firstly, there are the high maintenance costs of managing outdated systems over time—resources that can be devoted to other growth initiatives. Inefficiencies that result from slow response times can also reduce productivity. This is alongside poor integration with modern applications, which restricts scalability and adaptability to changing business requirements.  

Then there are security concerns, with Yugal Joshi, partner of IT services at Everest Group, pointing out that a robust cybersecurity posture is necessary to counter vulnerabilities of legacy systems. “As artificial intelligence scales, legacy systems will be more vulnerable to cyber threats,” he says. 

Modernising for the AI era

Bringing legacy modernisation back into sharp focus is the current era of artificial intelligence (AI). More than ever, this technology has presented companies with an opportunity to radically reshape the narrative around cumbersome legacy modernisation initiatives.  

For instance, Turner suggested that businesses tap into large language models (LLMs) to translate code into natural language, such that it can be summarised and presented to stakeholders. “That reduces the risk level of optimisation efforts because I can actually understand what I have—what the dependencies and impacts are—so I can turn code into specifications that I can use to build a more modern system,” he adds. 

And with AI gradually incorporated across myriad processes and applications, there is no better time for businesses to reimagine new technological foundations that are compatible with modern applications. Such an outlook will transform the developments in legacy modernisation.  

That’s why businesses should consider how to further automate and streamline existing systems, with AI at the core of this transformation. “This is a fantastic opportunity for organisations going through this optimisation process,” says Turner. “Businesses can plan for the next five, ten, fifteen years of transformation that we’re going to see, as AI comes into the mainstream and scales across applications.”  

Guiding modernisation with a clear business strategy 

Before setting the wheels of AI-enabled legacy modernisation in motion, businesses need to have a comprehensive understanding of their existing system and consider their strategic objectives.  

Rather than simply overhauling their legacy systems, Turner and Joshi agreed that these goals should be discussed and outlined. This can mean prioritising areas in the legacy system that are causing IT complexity and how they can be simplified to accelerate change.  

Another approach is seeing which aspects of business will benefit most from the new capabilities of AI and how this can guide legacy modernisation strategies. “With AI changing the way things are done within software engineering in general, as well as in modernisation, businesses, service partners, and their technology partners will need to figure out the right path forward,” says Joshi. 

Driving modernisation success is also about framing this change as a portfolio of opportunities. This includes considering whether any upgrades can be tied to key performance indicators, from driving revenue growth to reducing operational costs. 

“Think about tooling and automation. Think about factories. Think about standard methods. Think about how to balance systems that are being changed and migrated together, with ongoing business operations. Think about modernisation as building a capability over time, for the long term, where initial activity can realise savings, which can then fund that capability and accelerate those initiatives over time,” Turner explains. 

Improving business outcomes  

Several companies that have embarked on their AI-enabled legacy modernisation journey have seen results beyond just efficiency gains. One healthcare organisation has worked with Cognizant to streamline their membership platform that was running on multiple legacy systems, which has led to fragmented data silos.  

Through modernisation, the company has redesigned the application’s user interface, better integrate their data, and eliminate these silos, resulting in a tool that can offer customised recommendations to their users. “Having all of the data in place and being able to come up with good recommendations based on that data is really compelling to users,” says Turner. “We increased the registration rate for that application by 125 percent.” 

At the same time, Joshi shared that some companies have seen an increase in uptime of their operations—an outcome of a more resilient system. Another has reduced their partner onboarding process for an onboarding portal from three days to just a few hours.  

From improved customer experiences to reduced operational costs, legacy modernisation can yield benefits that can help businesses thrive in the age of AI.  

Find out how more about modernising your legacy systems in a landscape that’s engineered for AI with Cognizant.