Can AI Enable the Headless Enterprise?
Meet the Experts
Key Takeaways
⇨ The integration of multiple systems, including third-party applications alongside core SAP solutions, is becoming standard in enterprise landscapes, moving towards a composable ERP model that emphasizes flexibility and connectivity.
⇨ AI is increasingly viewed as a primary interface for business processes, prompting organizations to prepare their systems and data for future AI applications, potentially transforming traditional workflows and improving user interactions.
⇨ As companies anticipate the expansion of AI agents for task automation, establishing governance and a central repository for these agents will be crucial to manage interactions and ensure trustworthiness in AI-driven processes.
According to SAPinsider research, 52% of SAPinsiders now have an enterprise landscape that consists of core SAP solutions in combination with select third-party applications. Another 19% run a vendor-agnostic combination of best-of-breed solutions with select SAP applications, and 11% operate multiple ERP systems with select third-party applications. Only 18% report that they run predominantly SAP solutions.
This is a significant change over the last decade as many organizations move towards what has been called composable ERP—a flexible, customized environment consisting of different modules or components compared to a monolithic ERP system. Even those that are predominantly on SAP landscapes need to connect their SAP systems to non-SAP systems, making integration one of the most crucial parts of today’s business landscape.
Integration, Automation and the World of AI
One of the things that makes integration so important is that, for example, business processes do not end at the edge of the SAP landscape. They flow across systems and solutions, regardless of vendor. This makes connecting those systems a vital part of any business process, but especially for those with multi-vendor landscapes. This becomes even more important for those starting to add cloud-based solutions to their enterprise landscape.
Explore related questions
To understand more about this challenge, SAPinsider spoke with Boomi CEO Steve Lucas and Chief Product and Technology Officer Ed Macosky about how these changes play into what Boomi is doing with integration, automation, orchestration, and particularly AI agents. Boomi plays an important role for many SAP customers integrating third-party solutions. — Now, with Boomi’s newly announced Agent Designer and Agent Control Tower enabling users to create AI agents without writing code to meet their specific business needs, the company is making it easier than ever to drive automation and innovation within the SAP ecosystem. Innovating Beyond the System of Record
According to Lucas, SAP is the data platform or system of record for many organizations today. For example, something like an available to promise (ATP) scenario will always exist within SAP because it’s very deterministic. But Lucas regularly talks with customers that are looking to innovate once they get beyond the specific business process in SAP. This isn’t necessarily negative as the focus is on future innovation. However, customers are thinking about what they do in SAP and what might be handled by an AI agent in the future. This correlates with the rapid growth and deployment of AI agents across the enterprise landscape.
A potential future that Lucas foresees is that of the headless enterprise, an environment where the presentation layer is separated from the content repository or system of record. While this was difficult to bring about without AI, it is now much easier to envisage. Lucas said that Boomi has a significant role to play in this since they can tell the AI about the systems, context, and metadata. Moreover, he believes that this is all part of a move towards a future where users will no longer be directly interacting with many of these systems as those interactions will be handled by AI and AI agents.
For this to happen, Macosky stated that there will need to be standardization. This is something that many SAP customers went through years ago with SAP Business Warehouse and communication between systems and businesses. An organization like Boomi’s value in this is that it is working to integrate those standards and will also help bridge them. The Boomi Enterprise Platform can communicate between standards and, as the market determines the standard between agents, Boomi will be there to support that.
Macosky agrees that the headless enterprise is becoming more likely. In one example the customer may want a Boomi agent to reach out to an end user, collect an input, and then return the information to the workflow. Some of the customers that have had early access to Agent Designer over the last few months have created processes like this. As they augment their workflows, they aren’t going into the application but are using Boomi to connect to the user and get the information they need either through voice or a natural language input. Supporting this innovation beyond the system of record is something Boomi is already contributing to.
Governing Agents and Agent Interactions
Given this growth in AI, it seems obvious to suggest that companies will need some type of agent management platform. A year ago, there was little expectation that companies would need more than a handful of agents. Today, that governance is becoming crucial. In two years, Lucas suggested that companies might have swarms of agents that act based on user intent. While that isn’t happening today, it could easily be a future that is a couple of years away.
However, to make this possible Lucas indicated that companies would need a central repository of agents, showing where they are located, what they do, and what actions they can take. Beyond the first steps of governance that are being discussed today to make AI trustworthy, a directory of agents will be necessary to build that agentic swarm. Boomi is starting to build that directory so that whether organizations are running SAP, Boomi, or other third-party offerings, they will be able to more effectively manage and leverage AI.
What This Means for SAPinsiders
Today’s enterprise landscape is changing. While SAP solutions continue to be the system of record for SAPinsiders, there are now many third-party applications in the picture. Just as SAP is positioning Joule as the primary method of interaction for users, organizations are also looking at AI to function in the same way in business processes that span both SAP and non-SAP solutions. The ability to ask a written question rather than having to click through tiles or links can significantly simplify interaction for users and allow them to access information and make decisions swiftly.
What was possible with AI a year ago has already changed significantly and will continue to do so at a rapid pace. This time next year, organizations may be looking for ways to add AI capabilities to every business process. Some of these will require human interaction while others may simply call or orchestrate other agents to streamline and accelerate tasks.
While many SAPinsiders are still trying to determine their path forward as the end of mainstream maintenance quickly approaches, others are already putting in place the foundation for future AI success. Given these challenges, what should SAPinsiders do to be ready for the future?
- Start preparing your systems and data for future AI usage. While AI may not be part of your thinking today, the use of AI agents to streamline and accelerate processes and repetitive manual tasks is likely to become commonplace over the next year. This makes having harmonized data vital. The adage of “garbage in, garbage out” has never been more accurate than when it comes to AI. The more preparation work done today, the better prepared you will be for the future. From a systems perspective, it may simply be possible to connect systems to platforms in which AI can be leveraged, for example SAP BTP or the Boomi platform. Thus, Organizations must ensure that their systems are ready for that access.
- Start evaluating business cases where AI may prove advantageous. Most organizations likely have scenarios where adding AI to the mix will be beneficial. This could be anything from streamlining expense reporting to identifying suppliers from purchase orders. Starting to plan for these scenarios today will help make the data and system preparation process more efficient as there will be an understanding of where to focus.
- Understand the benefits of a future where AI is the primary interface to applications. Some organizations may be reticent to move away from traditional user interfaces, but it is important to fully understand the value of the headless enterprise. If this can be effectively implemented over the coming years, users may be able to see significant productivity gains and facilitate a greater focus on higher priority tasks. But organizations cannot take this step unless they are ready and understand the benefits it offers.
