Anantha Kondalraj
Head of Oracle Alliance
April 7, 2026
ERP transformation is being rewritten by AI. Not because of the ERP itself, but because intelligence is now expected to operate across the enterprise, rather than being confined to individual systems. Oracle’s recent innovations make this clear: AI is moving beyond features into cross-functional execution. Yet most ERP programs are not architected for this reality.
The traditional ERP core was built for process consistency, not intelligence, adaptability, or real-time decision-making at scale. AI is doing more than just enhancing these applications; it is reshaping how they operate and how decisions are made across them. The organizations that succeed will not be those that adopt AI fastest, but those that architect it correctly.
The Hidden Risk of Architecting AI the Wrong Way
Most organizations are introducing AI without evolving their architecture. Each platform embeds its own models. Each function builds its own automation. Each team defines its own governance.
First, this creates value. At scale, it creates risk:
- Fragmented decision-making slows execution and limits impact
- Inconsistent governance introduces exposure across departments
- Duplicated capabilities increase cost and complexity
- Operational instability emerges without production discipline
For CFOs, this raises questions about auditability and compliance. For CIOs, it creates challenges around resilience, cost control, and integration. For boards, the question shifts from “Do we have AI?” to “Can we trust it at scale?”
The issue is not model capability; it is architectural cohesion.
ERP Transformation Is Being Redefined
AI adoption is accelerating across every function. It is becoming part of how enterprise decisions are executed. However, enterprise reality extends beyond any single platform – most organizations operate across a complex landscape of applications, data, and legacy systems that must work together. As a result, the challenge is no longer embedding AI within applications, but enabling AI to operate consistently across them.
This is most visible in Oracle-led and multi-ERP environments, where organizations modernize core platforms while embedding AI across functions.
A new class of agentic applications in Fusion reflects this shift. Coordinated AI agents operate across finance, HR, supply chain, and customer processes, designed to reason, decide, and act within enterprise workflows. However, these capabilities remain confined within systems, not across them.
Why ERP Transformation Must Evolve
Most ERP programs are still designed around process standardization, data harmonization, and system consolidation. Decisions now span the enterprise:
- Revenue forecasting connects CRM/ERP
- Inventory decisions link supply chain and finance
- Workforce planning crosses HR and operations
- Risk and compliance extend across systems
If AI cannot operate across these domains, ERP becomes constrained by the architecture around it. Decision speed slows, control weakens, and value stalls.
Enterprise applications are shifting from systems of transaction to systems of outcome, where intelligence is embedded end-to-end, and decisions execute within the flow of work.
Hitachi’s Approach: Center for Architecture & AI (CAAI) as the Enterprise Intelligence Layer
Hitachi Digital Services builds, integrates, and runs platforms where AI operates as part of the architecture, not simply as an add-on. Our Center for Architecture & AI establishes AI as a shared operating layer across enterprise applications.
Intelligence is governed once and executed consistently across the digital core. It is designed for complex, hybrid environments including Oracle, SAP, Workday, Salesforce, and ServiceNow.
CAAI delivers four core capabilities:
- Architectural integration across legacy and cloud systems
- Governance by design through our R2O2 framework
- Enterprise orchestration across models, agents, and services
- Production-grade reliability through HARC for AI
Oracle’s introduction of an AI data platform as a semantic layer across enterprise data reinforces this direction. Scaling AI is not about embedding models in individual systems; it requires an architectural layer that enables intelligence to operate consistently across them. This is how AI moves from experimentation to engineered infrastructure.
Five Shifts Leaders Should Focus on Now
ERP remains central, but its role is evolving. It is becoming the execution engine for coordinated, intelligent decisions. This requires enterprise-grade governance, including security, auditability, and policy enforcement embedded directly into the platform. As AI becomes part of operational decision-making, this level of control is foundational.
Organizations moving from pilots to enterprise scale should focus on five shifts:
- Reframe AI as enterprise infrastructure
- Map the AI footprint to identify duplication and gaps
- Standardize governance early
- Enable cross-domain orchestration
- Engineer for production from day one
The goal is not more AI. It is AI that works consistently, at scale, under control.
The Strategic Choice
Organizations can continue modernizing systems and layering AI within them. Or they can establish a cross-enterprise operating layer that aligns intelligence, governance, and execution. Hitachi Digital Services helps organizations take that step by designing and running AI-led platforms that connect IT and OT, and are built for reliability, scale, and real-world execution.
As long-standing Oracle partners, we are experiencing this shift firsthand. The opportunity is not just to adopt AI, but to architect it for scale, control, and real business impact. The organizations that get this right will not just modernize ERP, they will redefine how their business operates.