Written by Konda — Paddy & Water Inc. / Strategy

SAP_Product_Roadmap_Complete_Guide_v2

SAP Product Roadmap: A Complete Guide

A Strategic Guide for CIOs and IT Department Managers 2025-2026

Expanded Edition: AI Future Outlook

July 2026

This report is based on publicly available information and systematically explains the current state and future roadmap of SAP products. It places particular emphasis on a broad and deep examination of the AI outlook from 2026 onward.

Chapter 1　Overview of SAP’s Product Strategy

SAP SE (headquartered in Walldorf, Germany) is the world’s largest enterprise resource planning (ERP) software company. As of fiscal year 2024, it serves more than 440,000 customers in over 150 countries, with global revenue of approximately EUR 35 billion. The core concept it has championed throughout the 2020s is the “Intelligent Enterprise,” which aims to organically combine ERP, data platforms, and AI to automate business processes and elevate decision-making. Then, at SAPPHIRE 2026 in May 2026, SAP further advanced this strategy by declaring the next phase: the “Autonomous Enterprise.”

1-1　Overall Structure of the Product Portfolio

• [ERP Core] S/4HANA Cloud Public Edition / Private Edition / On-Premise

• [Line-of-Business Applications] SAP Ariba (procurement), SAP SuccessFactors (HR), SAP Concur (travel and expense), SAP Customer Experience (CX)

• [Technology Foundation] SAP Business Technology Platform (BTP) — a PaaS providing integration, extension, AI, and data management

• [Industry Specialization] Industry Cloud (solutions for 25+ industries)

• [AI Foundation] SAP Business AI Platform (BAIP) — newly established in May 2026, a dedicated AI platform integrating BTP, BDC, and AI capabilities

1-2　2026 Key Themes

Theme	Overview	Related Products
Autonomous Enterprise Declaration	Establishes the “Autonomous Enterprise,” in which AI agents execute business processes autonomously without human instruction, as the future destination	Joule / S/4HANA / BTP
Business AI Platform (BAIP)	New AI platform brand integrating BTP, BDC, and Business Transformation Management	SAP BTP / BDC
Anthropic Claude Partnership	Adopts Claude as the primary reasoning engine for Joule agents, applied across 224 agents and 51 Joule Assistants	Joule / BAIP
Thorough Clean Core Adoption	Moves customizations to the BTP extension layer to ensure safe upgradability of the ERP core	S/4HANA / BTP
Renewal of the Data Foundation	Integrates Databricks, BigQuery, Snowflake, and Fabric into SAP Business Data Cloud (BDC)	BDC / Datasphere
Internalizing Sustainability	Embeds carbon costs into all business processes with Green Ledger as a standard S/4HANA feature	Green Ledger / S/4HANA

Chapter 2　SAP S/4HANA Roadmap

S/4HANA is SAP’s core ERP platform, built on the in-memory database SAP HANA and launched in 2015. As the successor to the legacy SAP ERP (ECC), it comprehensively covers all business areas including finance, procurement, manufacturing, sales, logistics, and human resources.

2-1　Overview of Deployment Models

Model	Characteristics	Primary Target Customers	Update Approach
Public Edition (Public Cloud)	Multi-tenant SaaS. Emphasizes standardization; deep customization not possible	Mid-size and small/medium enterprises; new SAP adopters	Two mandatory updates per year (May/November in 2026)
Private Edition (Private Cloud)	Single-tenant. Operated on hyperscalers (AWS/GCP/Azure). Delivered via RISE with SAP	Large enterprises and existing SAP users with complex business processes	2 to 4 times per year, at the customer’s own pace
On-Premise	Operated in the company’s own data center. Full control	Enterprises with strict regulatory and security requirements	One periodic release per year (Feature Package Stack)

2-2　Release Schedule (2025-2027)

Release	Preview Start	Go-Live	Key Focus
2026 FHY (First Half)	April 13, 2026	May 15, 2026	Generative AI integration, enhanced ESG reporting, financial automation
2026 SHY (Second Half)	October 12, 2026	November 13, 2026	Expansion of agentic AI, deepening of Industry Cloud
On-Premise S/4HANA 2025	Released October 2025	Mainstream maintenance through December 2030	Three FPS deliveries
On-Premise S/4HANA 2027	Scheduled for release October 2027	Mainstream maintenance through December 2032	Next-generation release standardizing AI and Clean Core

Chapter 3　Cloud Migration Programs: RISE, GROW, and ECC Maintenance Deadlines

3-1　RISE with SAP (Migration Bundle for Large Enterprises)

RISE with SAP is a bundled cloud service that supports existing SAP ECC/R/3 users — particularly large enterprises with complex customizations — in migrating to S/4HANA Cloud Private Edition. It is positioned not merely as a “cloud migration” but as a “business transformation.”

Component	Description
S/4HANA Cloud Private Edition	Single-tenant ERP operated and managed by SAP. Hyperscaler chosen from AWS/GCP/Azure
SAP BTP Starter Pack	Includes Integration Suite, Build Process Automation, and others in the initial configuration
Business Process Intelligence	Visualizes processes through process mining with SAP Signavio
Technical Migration Support	Provides ABAP code migration checks, the Data Migration Tool (DMLT), and the Readiness Check
Cloud Credit Scheme	Existing licenses can be converted into SAP Cloud Credits and applied toward RISE

3-2　GROW with SAP (For Mid-Size and Small/Medium Enterprises)

A program for mid-size and small/medium enterprises that emphasizes a multi-tenant environment and standardization, targeting go-live within 3 to 6 months. Customization flexibility is limited, but the cost structure is simple and subscription pricing is transparent. It grew rapidly from 2024 to 2025 and has become SAP’s primary engine for acquiring new customers.

3-3　SAP ECC Maintenance Deadline and Migration Pressure

Mainstream maintenance for SAP ERP (ECC 6.0) ends on December 31, 2027. As of the end of 2024, more than 60% of SAP ECC users had not yet completed their migration.

Option	Deadline	Cost Impact	Notes
Mainstream Maintenance	December 31, 2027	Standard maintenance fee	Standard support including new features and security patches
Extended Maintenance	December 31, 2030	+9% (vs. standard)	Covers EHP 6-8. Emergency fixes only; no new features
RISE with SAP Transition Option	December 31, 2033	Requires a RISE contract	Extended maintenance benefit for Business Suite 7
Third-Party Maintenance (e.g., Rimini Street)	Indefinite (contract-dependent)	Roughly 50% below standard	No official SAP support. Restrictions on AI functionality

Chapter 4　Clean Core Strategy

4-1　What Is Clean Core?

A design principle to “keep the ERP core in a state that can be safely upgraded through SAP’s release cycle.” It prevents erosion of the core by restricting custom development to “ABAP Cloud using only Released APIs” or “side-by-side extensions on BTP.” Clean Core is both a technical prerequisite for S/4HANA migration and an architectural requirement for Joule agents to function reliably (the premise for safety being that agents operate only through standardized APIs).

4-2　The Four Levels of Extensibility

Level	Definition	Upgrade Safety	Recommended Approach
Level A	Uses only Released APIs. ABAP Cloud (on-stack) or BTP (side-by-side)	◎ Fully safe	Standard for new development. The goal for all development
Level B	Uses officially documented SAP APIs (not Released APIs, but stable)	○ Generally safe	Permissible with governance approval; consider periodic migration to Level A
Level C	Accesses SAP internal objects (e.g., BAdI extensions)	△ Some risk	Register on a remediation roadmap; prioritize migration to Level A/B
Level D	Modifications, direct table writes, implicit enhancements	× Dangerous	Set a deprecation timeline and remediate promptly

4-3　The ABAP Cloud Development Model

A restricted subset of the ABAP language that permits access only to Released APIs. Support is progressing in both the SAP BTP ABAP Environment and on-premise/Private Edition environments. In Q4 2025, enhanced ABAP Custom Code Management and ABAP AI features were added. It works closely with Joule for Developers to support AI-assisted automatic code conversion from Level C/D to Level A.

Chapter 5　SAP Business Technology Platform (BTP)

5-1　The Role and Positioning of BTP

SAP BTP is a PaaS that functions as a “technology hub” connecting the ERP, LOB applications, external systems, and data sources. It runs on AWS, Google Cloud, and Microsoft Azure, and supports three runtimes: Cloud Foundry, Kyma (Kubernetes), and the ABAP Environment. From 2026 onward, its role as the execution foundation for AI agents will be further strengthened.

5-2　Key Service Groups

Service Group	Representative Services	Overview
Integration	SAP Integration Suite	An EiPaaS with 2,600+ connectors. Comprehensively covers API management, event mesh, EDI/B2B, and process integration
Extension	SAP Build (low-code/no-code)	Build Apps, Build Process Automation, Build Work Zone
AI Foundation	SAP AI Core / AI Foundation / GenAI Hub	GenAI Hub (multi-LLM access), Joule foundation, agent execution environment, Knowledge Graph
Data Management	SAP Datasphere	A unified semantic layer that integrates SAP and non-SAP data to provide an analytics foundation
Analytics	SAP Analytics Cloud (SAC)	Cloud analytics integrating BI, planning, and forecasting
Development Environment	SAP BTP ABAP Environment	Cloud-native ABAP runtime; the execution foundation for Clean Core extensions

5-3　Integration Suite Enhancements for 2026

• AI-Assisted Integration: A Copilot feature that generates and modifies iFlows using natural language

• Event-Driven Architecture: A real-time event broker via Advanced Event Mesh

• Pre-Built Business Content: Continued expansion of a library of pre-configured iFlows for SAP standard processes

BTP is evolving beyond simple “integration middleware” into the technical backbone supporting both the Clean Core architecture and the execution of AI agents.

Chapter 6　Business AI / Joule: Architecture, Agents, and ROI

6-1　The Evolutionary Trajectory of SAP Business AI

SAP’s AI strategy has evolved dramatically in a short period. Joule first appeared as a GenAI copilot in 2023, then evolved in 2024 into an embedded assistant within Fiori applications supporting transactions. Now, in 2026, SAP has redefined Joule from a standalone assistant into “a platform that integrates and coordinates 50+ domain-specialized Joule Assistants and 200+ specialized agents.” The biggest announcement at SAPPHIRE 2026 was the integration of this overall architecture as the “SAP Business AI Platform (BAIP),” positioning it as the technical foundation for the strategic vision of the “Autonomous Enterprise.”

Phase of Evolution	Timing	Key Capabilities	Characteristics
Phase 1: GenAI Copilot	Late 2023	Q&A, code generation, explanation	Centered on text generation; humans make decisions
Phase 2: AI Assistant Integration	2024	Embedded within Fiori apps; assists with transaction execution	Aware of process context
Phase 3: Agentic AI Platform	2025-2026	50+ Joule Assistants, 200+ agents	Autonomous end-to-end process execution
Phase 4: Autonomous Enterprise	2026 onward (target)	Autonomously executes entire business processes without human instruction	The stage at which AI “runs” the business

6-2　SAP AI Foundation: The Operating System for AI

SAP AI Foundation, announced by CEO Christian Klein at SAPPHIRE 2026, is the common AI foundation that underpins all Joule agents, the embedded AI within S/4HANA and SuccessFactors, and custom agents built with Joule Studio. SAP positions itself as “the operating system for business AI.”

The Three-Layer Structure of AI Foundation

• [Context Layer] The SAP Knowledge Graph (452,000 tables, 7.3 million data fields). Represents SAP’s domain knowledge in a machine-readable form, forming the foundation that lets agents understand business context

• [Build Layer] GenAI Hub (multi-LLM access), Joule Studio, Agent Builder, SAP Build. A set of tools for developing and managing agents, applications, and integration flows

• [Govern Layer] SAP AI Agent Hub, audit logs, access control, rollback capability. The governance foundation that provides visibility, control, and auditing of agent behavior

SAP Knowledge Graph

The Knowledge Graph is SAP AI Foundation’s greatest differentiator. It is not merely a metadata catalog, but an “enterprise knowledge network” that structures 452,000 tables and 7.3 million data fields extracted from S/4HANA together with their business context. By referencing this graph, Joule agents can accurately understand and act on business entities such as “accounts,” “cost centers,” “materials,” and “suppliers.” Since Q1 2026, a feature that automatically generates a Knowledge Graph from metadata within SAP HANA Cloud has reached general availability, reducing what previously took weeks to build down to minutes.

6-3　SAP’s Proprietary Foundation Models: SAP-RPT-1 and SAP-ABAP-1

Beyond providing access to external LLMs from OpenAI, Anthropic, Meta, Mistral, and others via GenAI Hub, SAP is also developing and deploying its own proprietary foundation models specialized for its business data.

SAP-RPT-1 (Relational Pre-trained Transformer)

SAP-RPT-1 is SAP’s proprietary foundation model specialized for structured, tabular business data. Whereas general-purpose LLMs tend to struggle with grasping relationships in structured data, SAP-RPT-1 can directly handle tasks such as “forecasting, prediction accuracy, anomaly detection, and optimization” in ERP, finance, manufacturing, and supply chain scenarios.

Comparison Axis	Description
Model Characteristics	A Relational Pre-trained Transformer specialized for prediction, forecast accuracy, and relational logic on tabular data
Performance Comparison	2x the predictive quality of narrow models (task-specific dedicated models); 3.5x the prediction accuracy of LLMs
Key Features	Retrieval-Augmented Prediction (RAP), column-level explainability (able to identify the business attributes underlying a prediction)
Use Cases	Demand forecasting, payment delay prediction, production schedule optimization, inventory anomaly detection, cash flow forecasting
Latest Version	SAP-RPT-1.5 (Q1 2026). Integration with the Knowledge Graph enables predictions that leverage business context

SAP-ABAP-1 (A Dedicated ABAP Code Model)

SAP-ABAP-1 is a code-specialized foundation model trained on 250 million lines of ABAP code, confirmed in the Q1 2026 SAP Business AI Release Highlights. Unlike general-purpose code LLMs, it generates code with an understanding of SAP-specific constructs as context — including BOPF, RAP (ABAP RESTful Application Programming Model), CDS annotations, BAdI implementations, OData service definitions, and Clean Core constraints. SAP states that its use reduces developer working time by 20-25%.

6-4　The Partnership with Anthropic Claude

At SAPPHIRE 2026, SAP announced a strategic partnership with Anthropic, revealing that it would adopt Claude as the primary reasoning and agent engine within the SAP Business AI Platform. This is one of the most important technology choices in SAP’s AI strategy.

Item	Description
Scope of Application	Functions as the reasoning layer for Joule and all Joule agents — 224 specialized agents and 51 Joule Assistants in total
Example Use Cases	Automating quarter-end book closing, responding to employee leave inquiries, automatically rerouting in-transit supplier orders
Integration Scope	Cross-system operations spanning S/4HANA, SAP SuccessFactors, and SAP Ariba
Technical Role	Multi-step reasoning, coordination across multiple agents, interpreting natural-language intent and formulating execution plans
Relationship to Other LLMs	OpenAI, Mistral, Meta Llama, Google Gemini, and others remain available via GenAI Hub. Claude is the default reasoning engine

6-5　Joule Studio 2.0 and Agent Builder

Joule Studio 2.0 is the latest version of the development environment that lets enterprises and partners build and manage their own Joule agents.

• Managed Agent Builder: Currently in Early Adopter Care, with GA planned for Q3 2026. A no-code GUI for defining an agent’s goals, permissions, triggers, and workflows

• Multi-Agent Workflows: Multiple agents collaborate across finance, SCM, and HR, resolving dependencies as they execute together

• Framework Support: Supports three frameworks — LangChain, LlamaIndex, and Microsoft AutoGen — enabling existing in-house AI assets to be integrated into the Joule ecosystem

• Integration with Business Data Cloud: Agents can reference structured and unstructured internal data in real time

• Governance and Auditing: All agent operations are recorded in an audit log. Approval workflows before action execution and rollback are both supported

6-6　SAP AI Agent Hub: Centralizing Agent Governance

SAP AI Agent Hub is a governance platform that lets organizations discover, monitor, and manage all AI agents — whether from SAP, partners, or built in-house — through a single console. It helps curb the risks associated with agent sprawl and provides visibility into who operated which agent, with what permissions, against which data. This capability is especially important for maintaining an auditable operational trail of AI agent actions to meet regulatory requirements such as J-SOX, GDPR, and ITAR.

6-7　Joule Agents by Business Domain and Concrete ROI

As of 2026, SAP offers more than 50 Joule Assistants and more than 200 specialized agents. Below are examples from key business domains along with published ROI figures.

Finance (FI/CO)

Agent / Function	Task	Published ROI / Impact
Cash Management Agent (GA Q1 2026)	Automatic analysis of daily bank statements, cash position preparation, automated reconciliation	An 80% reduction in cash position work time (official published figure)
Journal Posting Agent	Automatic posting of monthly journal entries, anomaly detection, closing checks	A 70-80% reduction in manual journal entry effort (reported by production sites)
Invoice Processing Agent	Automating the invoice receipt to matching to payment approval workflow	Cost reduced from $8 to $1.50 per invoice. At 100,000 invoices processed per year, an estimated savings equivalent to roughly JPY 65 million
Expense Report Agent	Automated review of expense reports, policy compliance checks	50-65% of expense reports processed with no human involvement

Supply Chain and Procurement (MM/PP/Ariba)

Agent / Function	Task	Published ROI / Impact
Production Planning Agent	Decisions on releasing production orders (automatic verification of material stock, capacity, and plans)	Autonomously handles 80% of routine PO release decisions, letting planners focus on the remaining 20% of exceptions
Supplier Onboarding Agent	Automating registration, screening, and master data creation for new suppliers	Substantially shorter onboarding lead time (specific figures depend on partner disclosures)
Procurement Agent	Automating purchase order creation, supplier evaluation, price comparison, and approval workflows	More than a 50% reduction in PO processing effort (pilot case figures)
Supply Chain Disruption Agent	Detecting delivery risk, proposing alternative suppliers, planning rerouting	A significant reduction in manual response time to mid-course changes

Human Resources (HCM / SuccessFactors)

Agent / Function	Task	Published ROI / Impact
HR Service Agent	Responding to employee inquiries about HR policy and handling leave requests	Automates more than 70% of routine HR inquiry responses
Payroll Agent (Autonomous HCM)	Detecting payroll anomalies, suggesting corrections, generating reports	Announced at SAPPHIRE; deployment cases expected from the second half of 2026
Hiring / Workforce Planning Agent	Managing the hiring workflow and supporting workforce planning	Shorter hiring lead time and improved accuracy of workforce plans
Upskilling Agent	Automating course recommendations and skills-gap analysis	Reduces analysis workload for L&D staff

6-8　Joule for Developers ABAP: AI for Developers

Joule for Developers ABAP (J4D ABAP) is SAP’s proprietary developer-focused AI tool for accelerating ABAP code development. As of Q1 2026, SAP announced that “the free promotional period for Joule for Developers has been extended through September 2026,” encouraging early adoption.

Key Features (as of 2026)

• Natural Language to ABAP Code Generation: Simply write “please create a CDS view” to generate a Clean Core-compliant CDS view

• Automatic Unit Test Generation: Automatically creates test classes from ABAP code, reducing the effort needed to improve coverage

• Code Review and Explanation: Automatically detects issues in existing code and proposes Clean Core-compliant alternatives

• Classic ABAP to ABAP Cloud Conversion: Detects the use of non-Released APIs and proposes real-time rewrites to Released APIs

• VS Code Extension (GA Q2 2026): AI-assisted ABAP development directly from VS Code via the ABAP MCP Server

2026 Roadmap (Officially Announced via SAP Community)

• Agentic AI Evolution: Moving from code completion to “agents that autonomously navigate a codebase to complete tasks”

• Continued Enhancement of SAP-ABAP-1: Automating the progression from code explanation (current) to code generation to code conversion (classic to ABAP Cloud)

• Deep Integration with SAP Business Application Studio (BAS): One-click ABAP refactoring within the IDE

Chapter 7　AI Future Outlook: The Path to the Autonomous Enterprise

This chapter explains SAP’s medium- to long-term AI vision. It centers on what was announced at SAPPHIRE 2026 and the direction SAP has officially indicated, organized in a way that IT decision-makers can apply to their medium- to long-term planning.

7-1　SAP’s “Five Key AI Themes for 2026”

In the SAP News Center, SAP presented the following five key themes that will define enterprise AI in 2026.

Theme	Description	SAP’s Corresponding Products
① The Rise of Specialized Foundation Models	In 2026, task-specialized models will surpass general-purpose LLMs on specific tasks. Models pre-trained on structured business data or specific industry data gain an advantage	SAP-RPT-1 / SAP-ABAP-1
② AI-Native Architecture	A shift from “adding AI to existing systems” to a design that “places AI at the core, with deterministic systems built on top of it.” Building infrastructure where AI interprets intent, retains context, and self-improves	SAP AI Foundation / BAIP
③ Generative UI Experiences	Instead of navigating screens and menus, users convey intent in natural language and AI carries out the operation. Say “arrange travel to visit my top customer lead” and Joule completes the booking	Joule / Autonomous Suite
④ AI Becomes Core Enterprise Infrastructure	AI functionality shifts from being an “optional add-on” to becoming standard infrastructure within ERP, HCM, and SCM. Failing to account for AI readiness in IT architecture design becomes a risk in itself	Embedded AI across all SAP products
⑤ The Institutionalization of Trust and Governance	As agent behavior proliferates across organizations, auditability, rollback, and cross-vendor governance become essential infrastructure	SAP AI Agent Hub

7-2　Autonomous Enterprise: SAP’s Ultimate Vision

SAP’s vision of the “Autonomous Enterprise” refers to a state in which AI agents autonomously execute business processes without step-by-step human instruction, freeing people to focus on handling exceptions and making strategic decisions. This vision rests on four principles.

• [Process-Aware] Agents act with an understanding of 7,000+ SAP business processes — grasping business context rather than merely executing tasks

• [Data-Grounded] Agents reference the SAP Knowledge Graph, with its 7.3 million fields, to accurately identify business entities before acting, minimizing the risk of hallucination

• [Governed by Design] Every agent operation is verified against permissions, roles, and access rights before execution. Audit logs are generated automatically, and agents themselves are aware of what they are permitted to do

• [Human-in-the-Loop by Default] Autonomous execution expands gradually. Today, nearly all significant actions include a human approval step, and full autonomy will advance as trust is built within each enterprise and business process

7-3　A Phased Scenario for Agentic AI Adoption (2026-2030)

Gartner forecasts that “by the end of 2026, more than 40% of enterprise applications will include task-specific AI agents (up from under 5% in 2025).” Combining SAP’s roadmap with broader industry trends suggests the following phased adoption scenario.

Stage	Approximate Timing	Typical State	What Enterprises Should Do
Stage 1: AI Assist	Through 2025 (present)	Joule answers questions and assists with on-screen operations. Humans handle all decision-making and execution	Begin Joule for Developers pilots; measure the ROI of developer productivity gains
Stage 2: AI Execute (Supervised)	2025-2026	Agents autonomously execute routine, high-frequency processes (with prior human approval). High-impact use cases such as the Cash Management Agent become mainstream	Move from PoC to production for in-house business agents built with Joule Studio; establish governance with AI Agent Hub
Stage 3: AI Orchestrate	2026-2028	Coordinated multi-agent collaboration handles end-to-end processing across multiple business processes. Humans focus on exceptions and strategy	Build out audit and compliance frameworks for AI operations; define AI ownership for business processes
Stage 4: Autonomous Enterprise	2028 and beyond	AI autonomously runs day-to-day operations. Executives and employees focus on judgment, creativity, and exception handling. The ERP itself becomes a “dynamic decision engine”	Transform organizational culture, skills, and governance as a long-term vision

7-4　AI Governance: A Framework CIOs Should Design Now

As the rollout of agentic AI accelerates, governance design is becoming directly tied to competitive advantage. SAP provides SAP AI Agent Hub, but deciding “what, who, and how” within governance remains the responsibility of the enterprise itself.

• Building an Agent Registry: Centrally manage all AI agents present within the organization — SAP-provided, third-party, and in-house built — via AI Agent Hub or a proprietary registry

• Defining a Permission Matrix: Clarify the types of actions agents may take, monetary thresholds, and the boundary between actions that require approval and those that do not (for example, POs under JPY 100,000 do not require approval; those above do)

• Audit Trail Retention Policy: Design the retention period and management framework for agent operation logs in consideration of J-SOX, GDPR, and ITAR

• Accountability and Remediation for AI Errors: Establish in advance the procedures for detecting, rolling back, and correcting agent misoperations

• Vetting Third-Party Agents: Establish a security review process for connecting external agents within Joule Studio

7-5　Changes in AI Cost Structure and Challenges Beyond 2027

As Forrester has pointed out, some of SAP’s AI functionality may see its “free runtime access period” end in 2027, and in many cases current customer budgets do not yet account for AI costs from 2027 onward. CIOs should scrutinize the following points now.

• Estimating licensing costs after the Joule for Developers free period ends (September 2026)

• Understanding the per-business-process cost of agent execution (number of API calls, LLM token consumption)

• Designing a unit-cost model for the BTP resource consumption (memory, CPU, storage) of in-house-built agents

• For companies using SAP third-party maintenance: evaluating the business cost of the constraint that they cannot make use of AI functionality such as Joule

Chapter 8　SAP Business Data Cloud (BDC)

8-1　A Shift in Data Strategy

SAP Business Data Cloud (BDC) is a new data foundation brand that integrates SAP Datasphere, SAP Analytics Cloud (SAC), and SAP Business Warehouse. It expands connectivity with external platforms — Databricks, Snowflake, Google BigQuery, and Microsoft Fabric — realizing the goal of “eliminating silos in SAP data while coexisting with existing analytics environments.” It also serves as a data source referenced by the AI Foundation’s reasoning engine, particularly SAP-RPT-1.

8-2　Integration Roadmap with External Data Platforms

Platform	Integration Method	Availability Status
Databricks	BDC Connect for Databricks. References the data catalog via Unity Catalog from the SAP side	GA October 2025
Google BigQuery	Via BDC Connect. Seamless sharing of SAP data products into BigQuery	GA planned for first half of 2026
Snowflake	Via BDC Connect. Bidirectional integration with existing data warehouses	GA planned for first half of 2026
Microsoft Fabric	BDC Connect support for Fabric, with Microsoft 365 integration also under consideration	Planned for Q3 2026

Chapter 9　SAP Industry Cloud

9-1　What Is Industry Cloud?

A set of solutions that extend industry-specific business processes without modifying the core of the standard SAP ERP. The structure builds industry-specific add-ons on top of the BTP extension layer, with S/4HANA as the core. Full compliance with Clean Core principles underpins the design philosophy. As of 2026, more than 25 industry-specific solutions are offered.

9-2　Key Solutions by Industry

Industry	Primary Solution	Characteristics / Features
Manufacturing (Discrete / Process)	SAP Digital Manufacturing (formerly MII/PCo)	Real-time production data connectivity, predictive maintenance, OEE optimization. Positioned as a cloud MES
Retail / Consumer Goods	SAP for Retail (Unified Commerce)	Unified inventory, customer loyalty, real-time analytics. Supports omnichannel (OMO) across e-commerce and physical stores
Life Sciences / Pharmaceuticals	SAP for Life Sciences (GxP-compliant)	Batch records, CSV compliance, FDA 21 CFR Part 11 compliance
Energy / Utilities	SAP for Utilities (IS-U)	Meter reading, billing, supply-demand management, smart meter support
Automotive	SAP for Automotive (EDI/OEM Integration)	Supplier management, EDI/JIT delivery, engineering change management
Financial Services	SAP for Banking / Insurance	Regulatory reporting (IFRS 17), risk management

Chapter 10　Sustainability and ESG: Green Ledger and Regulatory Compliance

10-1　SAP Green Ledger

SAP Green Ledger is SAP’s carbon accounting solution, which reached general availability in December 2024. It applies the same principle as a financial ledger to the concept of “managing carbon emissions as a double-entry ledger,” posting carbon emissions to S/4HANA in real time as transactions and events occur.

Function	Description
Real-Time Carbon Posting	Real-time posting linked to business transactions such as purchase orders, production orders, and shipments
Scope 1/2/3 Coverage	Covers direct emissions, purchased energy, and the entire value chain
Supplier Integration (Scope 3)	Collects and verifies supplier CO2 data through SAP Ariba
Regulatory Reporting	Report templates supporting EU CSRD, ISSB, EU ETS, and others
Embedding Carbon Costs into Operations	Supports decision-making that factors in carbon cost within procurement (MM), production planning (PP), and logistics (LE)

2026 is the year in which the first mandatory application cycle of the EU CSRD arrives for large enterprises. Under the 2026 roadmap, features that automatically embed carbon scores into procurement decisions, production planning, and logistics route optimization are scheduled to be released in stages.

Chapter 11　Strategic Recommendations for CIOs and IT Department Managers

11-1　Three Top-Priority Issues to Address

① Deciding a Response to ECC Maintenance End-of-Life (Deadline: December 2027)

It is advisable to complete S/4HANA licensing, fit/gap analysis, and deployment model selection (RISE or on-premise) within 2026. Extended maintenance (2030) and the RISE Transition Option (2033) merely extend the migration timeline — they are not means of avoiding the S/4HANA migration itself, and this point must be communicated clearly to executive leadership. It should also be emphasized that if a company adopts SAP third-party maintenance, the majority of AI functionality, including Joule, becomes unavailable.

② Defining Concrete AI Use Cases and Measuring ROI

The biggest failure pattern is “adopting an AI tool that produces no business impact.” A realistic approach is to start with off-the-shelf agents that have clear published ROI — such as the Cash Management Agent (80% time reduction) and the Invoice Processing Agent ($8 to $1.50 per invoice) — and then add company-specific use cases via Joule Studio. It is essential to simultaneously advance governance through AI Agent Hub and the design of audit logs for J-SOX compliance. For Joule for Developers, ROI measurement should be completed before the free period ends in September 2026.

③ Assessing Clean Core Migration and Visualizing Technical Debt

The premise for the safety of Joule agents is that they “operate only through Released APIs.” In environments where Level C/D customizations remain, the risk increases that an agent may manipulate data through an incorrect path. Use tools such as the SAP Custom Code Migration App (CCMX) and SAP Signavio process mining to quantify and prioritize technical debt.

11-2　Risk Management Matrix

Risk	Impact	Likelihood	Mitigation
Delayed response to the ECC maintenance deadline (2027)	High	High (over 60% not yet migrated)	Finalize the migration plan and establish a PMO within 2026
Agent misoperation due to inadequate AI governance	High	Medium (increasing as agent adoption expands)	Introduce AI Agent Hub; design a permission matrix and audit trail
Agentic AI malfunctioning due to unaddressed Clean Core issues	High	Medium	Conduct a CCMX assessment; formulate a roadmap for migration to Level A
Rising AI licensing costs from 2027 onward	Medium	High (per Forrester)	Model multiple cost scenarios within 2026 and budget accordingly
Overinflated expectations for BTP and AI not being met	Medium	Medium	Adopt a PoC-first approach; set ROI metrics in advance
Loss of AI functionality due to choosing third-party maintenance	High	Medium	Quantitatively explain to executive leadership the constraints on AI capability

11-3　Milestones for 2026-2027

Timing	Action Items
Q3 2026 (Now)	Decide on the ECC migration policy; select Joule/AI Foundation use cases; begin trialing Joule for Developers (taking advantage of the free period)
Q4 2026	Conduct PoCs and measure ROI for off-the-shelf agents such as the Cash Management Agent; complete the Clean Core assessment; formulate an AI Agent Hub adoption plan
Q1-Q2 2027	Formally begin the S/4HANA migration project; move in-house agents built with Joule Studio into production; establish the AI governance framework
Q3-Q4 2027	Complete cutover before ECC maintenance ends (12/31), or finalize extended maintenance. Full-scale operation of Stage 2 (AI Execute) AI agents

Conclusion

In 2026, SAP has clearly declared its transformation beyond the role of an ERP vendor into becoming an “AI foundation provider for enterprises.” At its core is SAP AI Foundation (the Knowledge Graph, SAP-RPT-1, and SAP-ABAP-1), along with more than 50 Joule Assistants and more than 200 agents powered by Anthropic Claude as their engine. However, when SAP speaks of the “Autonomous Enterprise” vision, it must not be forgotten that the greatest precondition is an environment where Clean Core and AI Foundation are properly in place. The practical adoption of AI agents is not a matter of technology, but a matter of foundational IT architecture readiness — ECC migration, Clean Core migration, and data foundation preparation. Before CIOs speak of an AI roadmap, firmly grasping the “groundwork” that must be laid, and accurately communicating the correct sequencing and ROI to executive leadership, is the single most critical IT priority for 2026-2027.

(This report was prepared based on publicly available information as of July 2026. SAP’s roadmap and release plans are subject to change without notice.)