SAP insider Magazine
The Philosophy of Openness
Boomi’s philosophy allows organiza- tions to standardize their data platform on non-SAP systems like Snowflake or AWS without being forced to make addi- tional SAP investments just to access their own data. “This means data extractions are faster and easier, while users regain full control over their data,” McLarty said. As organizations lay the groundwork for an AI-driven future, the ability to seamlessly integrate SAP data assets has become a prerequisite for survival. “Right now, organizations are laying the groundwork for the AI-driven fu- ture,” McLarty concluded. “First came generative AI. Now, organizations that can deploy AI agents with strong data management at their core will be ready for whatever AI innovation falls next.”
when, for the very first time, business leaders see fresh SAP data flowing into their systems and realize it’s actually working as promised,” he said. “What once took weeks of setup now happens in a few minutes or an hour.” He described a Day One scenario for an SAP Business Warehouse (BW) user to drive home this point: “You simply
Agentic AI is driving the next wave of innovation, but only if it has access to accurate SAP and non-SAP data.
For CIOs weighing their archi- tectural options, the choice often comes down to the eco- system philosophy: either sticking with SAP Integration Suite, which is strongest for SAP-to-SAP integration, or adopting a truly agnostic plat- form. McLarty argued for the latter, citing the reality that
select the SAP data entities you want to extract and specify where to land the data. The pipeline fetches historical data, replicates it, and then immediately switches to listening for incremental changes. It is a 360-degree view without the wait.” Fueling the Agentic AI Revolution Perhaps the most critical insight from McLarty involves the intersection of SAP data and artificial intelligence as the industry pivots from Generative AI to Agentic AI. However, the latter is only as intelligent as the data it can access. “Agentic AI is driving the next wave of innovation, but only if it has access to accurate SAP and non-SAP data and the ability to activate that data,” McLarty warned. He pointed to a recent MIT study that found 95% of corporate gen- erative AI pilots fail to deliver returns, mainly due to poor data quality and flawed integrations. McLarty offers a tangible example of this by illustrating how it would work for a supply chain decision platform. “By combining real-time SAP order data with external factors like live weather forecasts and satellite imagery of port delays, AI agents can proactively man- age risk,” he said. “With data that is instantly available and accurate, SAP customers can build AI agents that will base reasoning on the latest data points and act accordingly,” he added. This moves AI from a theo- retical science project to an operational reality that can handle invoice reconcili- ation, process sales orders, or flag low stock without human intervention.
modern enterprises are rarely 100% SAP. “SAP tools tend to be good when you stay within the SAP ecosystem,” he ob- served. “However, many organizations depend on non-SAP systems and data platforms as an integral part of their business process flows.”
What This Means For SAPinsiders
Break the ABAP dependency cycle. Boomi’s new Change Data Capture (CDC) capability operates at the application layer using a no-code approach. This effectively democratizes access to SAP data. Thus. IT teams can stop acting as bottlenecks for basic data work. By removing the need for specialized ABAP coding for data extraction, organizations can free up their most expensive technical resources to focus on complex process customization and migration tasks, such as the move to SAP S/4HANA, rather than writing brittle extractors. Real-time is now a prerequisite for agentic AI. Most SAP environments are accustomed to nightly batch jobs for SAP BW or data warehouses. While this latency was acceptable for yesterday’s BI dashboards, it can be lethal for tomorrow’s AI Agents. This is because autonomous AI agents cannot function on 24-hour-old data. By moving to an SAP Advanced Event Mesh in conjunction with SAP Integration Suite that pushes data changes in real time, SAPinsiders can transition their SAP data from a system of record to a system of intelligence. This allows them to deploy operational AI that can act on inventory or financial data in real time. Decoupling data with a heterogeneous approach. Matt McLarty argued for an open architecture. This approach enables organizations to use a neutral platform to extract SAP data into neutral data lakes such as Snowflake or Databricks, thereby avoiding vendor lock-in. This open approach allows them to standardize data outside of SAP. SAPinsiders get the best of both worlds: SAP remains their robust transaction engine. At the same time, the innovation layer can run on agile, non-SAP platforms without being penalized by difficult data access or the licensing complexities associated with indirect access.
let’s connect sapinsider.org 67
Made with FlippingBook - Share PDF online