Enabling innovation: Gen AI for the clean core and the future of software development

Enabling innovation: Gen AI for the clean core and the future of software development

Gianluca Simeone & Chiranth Ramaswamy

14 Apr 2025

• Facebook
• Twitter
• Linkedin

Developers building applications in just 20 minutes; that was one of the showcases at the SAP Sapphire event 2024.

The action may not have made major headlines at that time, but it was significant in pointing to the future of software development - and the two key factors shaping today's developer environment:

• With constant change defining the modern business landscape, the rapid pace of innovation must align with emerging opportunities.
• Progress also demands a rethinking of how new apps are delivered, with the emphasis placed on agility and speed to market.

Extending business capabilities

First, let's talk about SAP, and how its clean core strategy fits into this narrative.

Clean core is an approach that combats the limits placed on innovation by year-on-year customization of an organization's SAP estate. Such change is of course inevitable as new code and workflows are added to the core SAP product, but the end result has seen restrictions placed on how an SAP system evolves.

With the clean core approach, SAP BTP (Business Technology Platform) supports application development through a suite of tools and services designed to extend SAP applications without modifying the core:

• Extensibility: SAP BTP allows developers to create extensions that add to standard SAP software
• SAP Build Apps: a no-code platform that enables rapid application development without altering the core functionality of SAP S/4HANA
• Integration services: these ensure smooth communication between extensions and the core system (as well as between different SAP and non-SAP solutions).

The application of Gen AI

Against this backdrop of clean core and application development, Gen AI models are being trained to follow processes and pre-defined rules to help developers:

• Summarize all conversations that set out a project's parameters
• Understand where certain bits of code will need to sit
• Create an architected technical design following clean core principles
• Generate test data to streamline debugging and analysis.

What's more, these actions have become increasingly automated and surrounded by best practice recommendations. As a result, Gen AI can be tasked with understanding the required functionality, the SAP rules in play, and providing guidance for whether the development should occur inside the core or side by side with BTP. It can also suggest whether the functionality should be built using ABAP or Java, write the functional specification, and even generate code.

Supporting the clean core journey

Gen AI is therefore helping transform the application development process. Yet, for many organizations, the complexities involved in maintaining a clean core is proving too great a hurdle to overcome. Unsurprisingly though, Gen AI can also help here:

• The solution can be used to evaluate customizations within the existing core, analyze code, detail dependencies, and show the impact if removed - alongside the ideal path for moving critical code into the core.
• It can also look at established processes and create test cases for moving to the cloud, thereby bringing a higher level of confidence that any change will not compromise core business processes (then convert these test cases into a technical design).
• From a data quality perspective, Gen AI can also be used to generate reports evaluating the state of a data environment (including mappings, etc.), before generating scripts to monitor ongoing quality on a regular basis.

Inspiring greater creativity

What's therefore evident is that Gen AI can support the journey to the clean core. Hence why the technology's adoption is being primarily driven by the need to stimulate this type of innovation. Indeed, a recent report from Capgemini Research Institute - Generative AI in software engineering - suggests that enabling greater creativity and innovation was by far the biggest driver of adoption, followed by software quality and developer productivity.

Yet this last point, developer productivity, is just as important. Across the software development life cycle, a multitude of performance advantages can be achieved when the more mundane tasks (handwriting code from scratch, etc.) can be effectively automated. This leaves developers free to focus on more creative outputs, architecting solutions, and validating code rather than time-consuming code development.

Training will of course be an ongoing process for leveraging Gen AI to its fullest extent, particularly in areas such as prompt engineering. However, once user acceptance increases, a significant improvement in productivity can certainly be expected.

Final thoughts

Gen AI has the potential to become a developer's greatest assistant when creating new apps and undertaking the journey to the clean core. Across every step, from analysis and documentation through to code writing and final testing, the technology acts as a highly efficient assistant, capable of processing intricate details and adhering to precise, consistently applied rules. For today's developers, this opens up a world of productivity benefits; and, for those businesses that recognize the value of a clean core but are intimidated by the work involved to get there, Gen AI represents an ability to strip out much of the complexity involved - as well as the risk.