As enterprises embrace AI and modernize their SAP systems, a new book suggests that connecting business processes, architecture, data, and human decision-making is key to successful transformation.
Enterprise transformation has never suffered from a shortage of technology. Businesses the world over have spent decades investing in new applications, modernizing infrastructure, and implementing systems designed to make business operations faster, smarter, and more efficient.
However, connecting those investments to measurable business outcomes remains an ongoing challenge. As artificial intelligence introduces daily innovations and opportunities for automation and decision-making, the stakes couldn’t be higher.
For Kumar Piraviperumal, semiconductor industry leader in the IBM HiTech practice, addressing these challenges begins not with technology, but with understanding how a business actually operates. “My belief is process is the base of all transformation,” he said during a recent SAP PRESS book launch webcast.
That primary belief is central to his new book, Integrating Architecture, Processes, and AI on SAP BTP, which explores how organizations can connect business processes, enterprise architecture, data, AI, and employee adoption into meaningful transformation.
Instead of categorizing these capabilities as separate technology initiatives, Piraviperumal believes organizations need a connected foundation that carries business intent from process design through application development, deployment, and ongoing governance.
Turning transformation theory into practice
Piraviperumal’s book is based on an experience he had working on a transformation program for an $8 billion media company operating across 47 countries.
The organization ran complex business processes, and his team was among the early adopters attempting to connect SAP Signavio for process modeling with SAP LeanIX for enterprise architecture.
The work proved challenging. As the team encountered integration problems, missing information, and inconsistent approaches, Piraviperumal documented what worked, what failed, and how to standardize the process.
“The motivation was basically trying to create a practical book for everybody,” he explained.
The end result is a guide intended for practitioners who need to translate business requirements into functioning enterprise applications.
Its central theme is that transformation requires more than documentation or architectural diagrams. Organizations must create continuity between the processes, systems and the people responsible for using them.
Connecting the enterprise from end to end
Piraviperumal believes enterprise transformation is a progression that begins with business strategy and extends through value streams, capabilities, processes, resources, data, and technology in the following ways:
- SAP Signavio provides the process foundation, enabling organizations to analyze how work is performed and identify opportunities for improvement using operational data.
- SAP LeanIX connects those processes to the underlying application architecture, while SAP Business Technology Platform (SAP BTP) provides capabilities for integration, development, and AI-enabled execution.
- WalkMe adds another critical dimension: helping employees adopt new processes and applications.
This begs the question: How do you make these different technologies work together without losing the connections between business requirements and technical execution?
One method is through identifiers. A process modeled in SAP Signavio may have a different identifier in SAP LeanIX and another in WalkMe, making it difficult to maintain a consistent view across systems.
Piraviperumal’s team created identifier consistency by creating a common data structure that maps identifiers and connects them to the applications being developed.
This “canonical data spine” strategy establishes traceability across the transformation life cycle, which becomes extremely valuable when something goes wrong.
A defect in a business application, for example, might originate in the software itself, the original process design, or a business process that has become outdated. Properly managing those connections helps organizations identify the source of a problem rather than treating every issue as an isolated technical failure.
“Traceability is one thing which I keep stressing in this book,” said Piraviperumal.
Employee adoption before deployment
Piraviperumal emphasizes that organizations should consider how employees will interact with new systems before development begins, rather than treating change management as a final implementation activity.
His book underscores the importance of digital adoption concepts before moving into the technical build.
Using WalkMe, organizations can connect employee guidance and adoption information to the business processes being transformed. The objective is to understand how new workflows will affect employees, anticipate friction, and improve the likelihood that technology investments deliver their intended value.
“Business is not interested in SAP,” said Piraviperumal. “They are interested in outcomes.”
From process models to AI-generated applications
One of the book’s most practical themes is the opportunity to use existing business process models as inputs for AI-assisted application development.
During the webcast, Piraviperumal demonstrated how a process designed in SAP Signavio could help generate an application for SAP Analytics Cloud, using a financial close-to-rolling-forecast process.
By exporting the approved process model and combining it with architectural information from SAP LeanIX, his team used AI capabilities to identify process issues, generate configuration information, and produce code supporting application development.
The demonstration also incorporated IBM watsonx and other tools developed by Cognitus.
Instead of relying on an open-ended natural language prompt, the approach instead relies on an existing business process model as the foundation for generating the application. This strategy gives developers a structured starting point while preserving connections to the original business requirements.
It also illustrates how AI reduces manual effort involved in translating process documentation into functional specifications, technical configurations, and executable applications.
However, Piraviperumal stressed that AI-assisted development does not eliminate the need for developers, configurators, or business validation.
Keeping humans in control of AI
For all of the cautionary attention surrounding autonomous AI, Piraviperumal believes governance and human oversight remain essential.
His transformation framework recognizes three broad constraints: economics, technology, and social adoption. Even when a solution is technically feasible, organizations must consider its cost and whether employees will accept it.
AI adds another layer of complexity because automated systems must operate within clearly defined boundaries.
Piraviperumal dedicates the book’s final chapters to operational governance, security controls, and audit requirements, arguing that these capabilities are fundamental to trustworthy deployment.
He also describes incorporating human validation checkpoints into the code examples included in the book.
“AI can only do so much,” he said. “AI cannot make decisions. It can help you with options to make a decision but not on decisions itself.”
A more connected future for enterprise transformation
The book’s broader message underscores the importance of successful transformation and how it depends on preserving the relationship between business intent and technology execution.
Process modeling, enterprise architecture, application development, AI, and employee adoption cannot operate effectively as disconnected disciplines. It must support a continuous, governed system that enables organizations to understand what they are changing, why they are changing it, and whether the results deliver value.
For practitioners, the opportunity is to use that connected foundation to make enterprise transformation more repeatable, traceable, and responsive to changing business needs.
And even as AI becomes more capable of generating applications and supporting business decisions, Piraviperumal sees human judgment as an enduring part of that equation.
“Always remember that human control is still important even in this AI world,” he said.
