Contents
Where AI Creates Value on Top of SAP
Architectural Choices That Matter
Mapping AI Use Cases to the Best-Fit Platform
Turning Strategy into Action: Start Building ML on Harmonized Data in the Cloud
AI capabilities in SAP-centric enterprises are no longer delivered by a single platform. Instead, they emerge from a hybrid ecosystem that combines native SAP innovations with best-of-breed cloud AI services. SAP continues to embed intelligence directly into SAP S/4HANA and the Business Technology Platform, accelerating time-to-value for standard business processes such as finance, logistics, and procurement.
At the same time, advanced and large-scale AI use cases increasingly rely on non-SAP platforms, particularly hyperscalers, to take advantage of their scalability, compute power, and rapidly evolving AI services. The most effective AI strategies do not focus on choosing SAP or non-SAP solutions, but on orchestrating both—ensuring SAP remains the trusted system of record while cloud AI platforms provide the flexibility and scale required to drive innovation.
As artificial intelligence continues to transform how businesses operate, SAP customers now have access to a broad spectrum of AI capabilities. From automating financial processes to optimizing supply chains and enhancing customer interactions, AI-driven solutions can unlock significant business value. However, selecting the right AI use cases and platforms remains a challenge, requiring a clear understanding of business priorities, data readiness, and platform maturity.
Where AI Creates Value on Top of SAP
SAP customers should start their AI journey by identifying use cases that clearly align with business objectives, data readiness, and the AI capabilities available within their SAP landscape. The first step is to assess whether prebuilt AI features in SAP S/4HANA, SAP Analytics Cloud (SAC), SAP Concur, SAP SuccessFactors, or other SAP applications already address these needs, or whether custom AI models are required. As part of this assessment, organizations should evaluate whether SAP AI capabilities on SAP BTP are sufficient to implement the identified use cases at the required scale and maturity.
In recent years, Generative AI and large language models (LLMs) have received significant attention. While these technologies are important, they represent only one component of a much broader AI landscape. Many AI and machine learning techniques have been successfully applied across industries for years, delivering proven, measurable business value.
Common industry use cases include customer churn analysis, where organizations—particularly in the telecommunications sector—classify customers based on their likelihood of switching to a competitor. Another widely adopted scenario is market and price forecasting, where regression models are used to predict demand trends or anticipate price movements for specific goods.
Machine learning is also frequently used for customer segmentation, grouping customers based on behavior, demographics, or preferences. This enables marketing teams to design more targeted and effective engagement strategies. In addition, recommendation systems leverage historical sales and interaction data to deliver personalized product and service suggestions.
These examples illustrate just a subset of the many ways machine learning is applied in real-world industry scenarios today. Beyond the current GenAI hype, both traditional and advanced ML techniques remain essential for driving operational efficiency, revenue growth, and sustainable competitive advantage in SAP-centric enterprises.
Architectural Choices That Matter
Successful AI architectures in SAP-centric enterprises rely on seamlessly integrating SAP data platforms with non-SAP sources, while keeping SAP as the trusted system of record. With capabilities such as SAP Integration Suite, SAP Business Technology Platform (BTP), and SAP Business Data Cloud (BDC), organizations can harmonize transactional and analytical data within a secure, governed AI architecture. This harmonized foundation is critical for AI agents—like SAP Joule—to operate with accurate business context, ensuring trust, compliance, and explainability.
At its core, SAP acts as the integration and orchestration layer for enterprise AI. By connecting systems and standardizing business data, SAP enables organizations to scale AI beyond isolated projects. While custom AI on BTP can deliver strategic differentiation, it often requires deep expertise, high-quality data, and longer timelines, which can delay ROI. Embedded AI capabilities within SAP applications, by contrast, provide faster time-to-value for standardized business scenarios.
Most enterprises adopt a hybrid AI strategy: embedded AI delivers quick wins, custom AI on BTP addresses strategic use cases, and hyperscalers provide the scalable compute, advanced models, and innovation velocity required for complex, large-scale workloads. Cloud-native platforms and AI services on AWS, Azure, or Google Cloud enable machine learning, deep learning, and Generative AI applications by securely consuming harmonized SAP and non-SAP data via APIs, pipelines, or streaming technologies. These layers work together to move organizations from tactical automation to end-to-end, AI-driven business processes, unlocking broader value and sustaining trust in AI outcomes.
The following diagram illustrates an example data architecture, showing how SAP S/4HANA and additional data sources support operational reporting and analytics, while hyperscalers enable advanced analytics, AI applications, and API-driven consumption.
Useful Links
The role of SAP BDC
SAP Business Data Cloud (BDC) is SAP’s strategic foundation for enterprise data, designed to enable a unified data fabric architecture across SAP and non-SAP landscapes. It consolidates SAP’s core data, analytics, and AI capabilities into a coherent platform that provides trusted business context for analytics and AI initiatives. At its core, BDC brings together:
- SAP Datasphere for business semantics, data modeling, and harmonization
- SAP Analytics Cloud (SAC) for analytics, planning, and BI enriched with Joule AI
- SAP Business Warehouse (BW) Private Cloud Edition for BW modernization
- SAP Databricks for advanced data engineering and AI/ML workloads
- AI Foundation as the AI operating layer, enriched with a business-aware Knowledge Graph
Importantly, BDC is modular by design. Each component remains a standalone product, allowing organizations to adopt only the capabilities they need based on their current architecture, data maturity, and business priorities. For example, a company may leverage SAP Datasphere to harmonize S/4HANA and non-SAP data without deploying SAP Analytics Cloud or SAP Databricks.
SAP BDC is also built to coexist with hyperscaler-native data platforms such as Google BigQuery, Snowflake, Databricks, or Microsoft Azure Synapse. In many architectures, SAP Datasphere provides governance, business semantics, and trusted data access, while large-scale analytics and AI/ML workloads run directly on hyperscalers. In this role, SAP BDC acts as the data backbone that enables a multi-platform AI strategy—providing structure, consistency, and trust without creating platform lock-in.
Useful Links
Mapping AI Use Cases to the Best-Fit Platform
Mapping AI use cases to the right platform starts with understanding their complexity and business impact. Standard, process-driven scenarios are best addressed with embedded AI in SAP applications, delivering fast time-to-value with minimal overhead. More advanced predictive and optimization use cases benefit from SAP Analytics Cloud or custom AI on SAP BTP, where governance and SAP process integration remain essential. Highly differentiated or compute-intensive use cases are best executed on hyperscalers, combining SAP’s trusted data foundation with cloud-scale AI capabilities.
| Use Cases | AI Requirement | Best-Fit AI Platform |
| Predictive Accounting Cash Flow Forecasting (short-term) Automated Invoice Processing Smart GR/IR Reconciliation | Standard AI embedded in business processes | SAP S/4HANA Embedded AI (incl. SAP Business AI, Joule capabilities where applicable) |
| Sales & Demand Forecasting Anomaly Detection (financial / operational) Predictive Maintenance Financial Planning & Analysis (FP&A) | Predictive analytics & forecasting | SAP Analytics Cloud (SAC) with AI + Hyperscaler ML services (for advanced models or scale) |
| AI-Powered Document Processing Intelligent Chatbots (SAP-context aware) ML-Based Risk Scoring & Classification Manufacturing Quality Control IoT Anomaly Detection Regulatory Compliance Monitoring AI-Driven HR Analytics | Custom AI models embedded in SAP workflows | SAP AI Core & AI Foundation (BTP), SAP Business Data Cloud (BDC) + Hyperscalers |
| Image & Video Recognition Advanced NLP & GenAI (LLMs) Large-Scale Fraud Detection Market & Sentiment Analysis | Large-scale, compute-intensive AI | Hyperscalers (AWS, Azure, GCP) |
Useful Links
Turning Strategy into Action: Start Building ML on Harmonized Data in the Cloud
Once SAP data is harmonized, governed, and integrated across the enterprise, organizations are well-positioned to move from AI strategy to execution. Hyperscalers represent the natural next layer to operationalize this foundation, providing the scale, elasticity, and managed AI services required to turn trusted data into actionable intelligence. Their cloud-native data platforms enable organizations to combine SAP data with non-SAP and external sources, experiment rapidly, train models efficiently, and deploy machine learning applications that deliver measurable business outcomes.
Rather than beginning with complex, high-risk initiatives, successful organizations focus on pragmatic ML use cases that build confidence and demonstrate value early. Common starting points include forecasting, anomaly detection, classification, recommendations, and optimization models. Hyperscalers reduce the barrier to entry through managed services, prebuilt algorithms, and integrated MLOps capabilities, allowing teams to focus on business outcomes instead of infrastructure and platform complexity.
The most effective enterprises do not wait for a “perfect” AI architecture. They activate harmonized datasets in the cloud, iterate quickly, and scale what works. By combining SAP’s strengths in data integrity, governance, and business context with the innovation speed and scalability of cloud AI platforms, organizations can accelerate their AI journey—moving from insight to action, and from experimentation to sustained competitive advantage.
A 90-Day Roadmap to Operational ML on GCP
Within 90 days, organizations can move from harmonized data to operational machine learning on Google Cloud (or another hyperscaler) through a phased approach:
- Phase 1 – Enable (Days 0–30): Establish secure data access and governance by exposing selected SAP datasets to BigQuery, ensuring alignment with enterprise security and compliance requirements.
- Phase 2 – Build (Days 30–60): Develop and validate ML models using BigQuery ML for rapid, analytics-driven use cases or Vertex AI for more advanced predictive and optimization scenarios.
- Phase 3 – Operationalize (Days 60–90): Deploy models, integrate predictions back into business processes, and establish monitoring and feedback loops to support continuous improvement.
This approach ensures SAP remains the trusted system of record, while GCP provides the scale and innovation required to deliver AI value quickly, pragmatically, and at enterprise scale.