Beyond the S/4HANA deadline: How to prepare SAP systems for scalable AI

Yes. The stakes are a lot higher. If there’s a problem when you’re transforming the financing and controlling processes in a company, it remains an internal issue. It may cost more money or time, but the problem stays inside the company.

Once SAP moves beyond finance, it changes how people experience the business. Issues don’t just impact your company. They affect the relationships you have with your customers and the relationships your customers have with their own customers and employees. Since you’re touching the heart and core of the business, you must be even more deliberate and get things right from day one.

AI only works if the foundation is clean. It’s very difficult to use AI in a responsible and scalable way if processes aren’t standardized and the core is overloaded with custom code. That’s why standardization is so important. It’s what allows you to move AI and agents beyond isolated use cases and really support the business.

Data quality is also essential. AI is only as good as the data behind it, so moving everything without discipline doesn’t help. You should migrate only the data that makes sense. Targeted migration eliminates non-essential data, which gives you a cleaner foundation for AI and better insights and helps manage costs in subscription models.

Absolutely. Many companies are on their second SAP transformation, so they’ve seen where things get complicated and what slows a program down. These experiences give them the comfort and confidence they need to move faster.

Expectations are also changing. Customers are looking for less lifting and shifting and more original thinking. They want a simpler core and fewer extras that don’t provide real value. And they’re searching for partners to help them on their journey, not just sell them products or services.

Kyndryl itself is a good example. As part of our ongoing transformation, we set out to modernize two SAP applications that contained more than 33,000 lines of custom code. We used Nova Intelligence’s AI solution and our Clean Field approach to simplify the core, remove direct table access and align with SAP best practices from the start.

The results were significant. The transformation took just 10 days and required 75% less manual effort. It also cut costs by 50% and, most importantly, created a cleaner foundation for AI and future innovation.

In the end, the success of SAP transformation isn’t defined by meeting a deadline but by how well the organization is prepared for what follows. As AI and agentic become part of everyday operations, choices about what to simplify, what to preserve and what to standardize made during modernization will shape the next phase of SAP and the company’s ability to adapt over time.