For legacy software, AI will drive several distinct trends, but not all of them center on AI. As a co-founder of a business specializing in legacy platform modernization, here are four key themes to watch over the course of 2025.
4 Ways AI Will Affect Legacy Software Platforms
- Businesses will use Clean Core to migrate their custom code out of SAP.
- AIOps will continue to merge with IT operations.
- Organizations will finally catch their breath and adapt for AI.
- We’ll see the second wave of data modernization.
1. Organizations Will Fully Embrace SAP Clean Core
One of the most significant trends in legacy software in recent years has been Systems, Applications and Products in Data Processing’s Clean Core initiative. The goal of Clean Core is to migrate custom code that businesses write themselves out of the SAP platform.
Embracing Clean Core is important for many reasons, but chief among them will be the ability of SAP customers to take full advantage of AI. SAP has indicated that it will steadily roll out new AI features and capabilities in the coming years, and each time they appear, users will need to upgrade to the latest SAP version to use the new AI offerings. When custom code is embedded in SAP, however, it complicates upgrades.
Clean Core ensures that organizations can access SAP’s latest AI capabilities without their custom code preventing or slowing down successful upgrades.
The Clean Core initiative has been underway for several years, but to date, many of the businesses I work with have struggled to migrate their custom code out of SAP. This is understandable, because it’s a complex process with each business facing a unique set of challenges.
But I foresee 2025 as the year when organizations finally start to address the big Clean Core migration hurdles they’ve been facing. In short order, Clean Core will become the norm of the SAP world.
2. IT Operations Will Keep Implementing AI Solutions
Even before widespread generative AI technology, folks were talking a lot about AIOps: using AI to automate and scale complex IT processes (like identifying the root cause of a server failure or provisioning user access rights on demand).
But now that generative AI has entered the picture, IT tool vendors have access to a wide range of new types of AI-powered use cases. They can, for example, use AI to help customers formulate and execute plans for migrating from legacy platforms to the cloud or implementing initiatives like Clean Core.
Since these processes involve parsing large volumes of data and making complex decisions, they are tedious and time-consuming to perform manually. AI can significantly streamline aspects of these operations. Although, don’t expect AI to eliminate the need to keep skilled humans in the loop when managing complex, risky operations like platform migration.
3. Business Processes Will Adapt for the AI Age
Beyond the IT space and certain other niches, businesses adopting AI has been relatively limited. Generative AI has slipped into what Gartner calls the trough of disillusionment.
I think the real challenge holding back use cases for AI in business is that business processes have not yet caught up with AI. To take full advantage of AI, companies need to do more than simply place AI tools in the hands of employees; they need to rethink the way employees work, too.
For example, if AI reduces the time it takes to complete a certain task by 90 percent, you need to reformulate your business process so that employees can efficiently repurpose all the time they save. Otherwise, AI simply shortens the duration of some workflows without creating real value.
It will take more than a year for businesses to make this pivot fully, but I suspect that 2025 will be the year when they’ll start to make the major process changes necessary to enable AI use cases that extend beyond the simple chatbots and workflow automation we’ve seen so far.
4. The Second Wave of Data Modernization Is Coming
Data modernization (the process of transforming data in ways that make it as useful as possible for a business) was a hot topic about a decade ago. Back then, big data was a big buzzword, and companies were investing in a new breed of data analytics, reporting and business intelligence solutions.
But given the proliferation of interest in AI, and the central role that high-quality data plays powering AI tools and services, I suspect we’re in the early stages of second-wave data modernization. Whereas early data modernization initiatives centered on helping companies take stock of their data and use it for purposes like predictive analytics, the new wave of data modernization focuses on preparing data for use by AI.
Doing so entails enhancing data quality like never before and working with new types of data, especially unstructured data, like documents or audio transcripts. These types of data are not as useful as structured data (like names and addresses in databases) for analytics, but they are key for tasks such as training chatbots.
