From ERP to ERX: How AI Is Turning Planning into Intelligent Execution

Enterprises have not necessarily lost faith in ERP modernisation. What they are questioning is the old transformation model: spend heavily, disrupt operations and wait years to find out whether the promised business value actually arrives. In an interaction with DQ Channels, Phil Lewis, Senior VP,

Enterprises have not necessarily lost faith in ERP modernisation. What they are questioning is the old transformation model: spend heavily, disrupt operations and wait years to find out whether the promised business value actually arrives. In an interaction with DQ Channels, Phil Lewis, Senior VP, Solution Consulting International (EMEA & APJ), Infor, discussed how ERP modernisation is shifting towards phased delivery, measurable outcomes and industry-specific technology, while AI is pushing ERP beyond its traditional transactional role towards Enterprise Resource Execution, or ERX.

Modernisation needs measurable outcomes

The big-bang ERP approach is increasingly difficult to justify when businesses cannot wait several years for returns. Instead of beginning with the question of which technology to deploy, organisations need to identify what they want to improve over the next six or 12 months.

That changes the structure of transformation.

Infor's approach, described as value mapping, identifies the financial and operational improvements expected from a project before implementation begins. Business priorities are ranked, less critical requirements can move into later phases and measurable KPIs provide checkpoints along the way.

The underlying principle can be reduced to three steps: identify the outcome, deliver it and measure it. This shifts ERP modernisation from a technology milestone to a business performance exercise. Each phase has to demonstrate value before the organisation moves further along the roadmap.

It also addresses one of the biggest risks in ERP transformation: moving faster than the organisation can handle. Enterprises need to modernise at the speed the business can absorb, rather than simply at the speed technology allows.

For systems sitting at the operational core of an enterprise, this distinction matters. Disruption can affect production, revenue, supply chains and customer delivery. Business continuity, therefore, becomes as much a change management issue as a technical one.

Infor's approach is to build Cloud solutions around specific industries and even micro-vertical requirements. A bakery, for example, does not operate in exactly the same way as a cheese manufacturer, despite both belonging to the food and beverage sector.

This industry-specific foundation can reduce the amount of customisation required during implementation. The company estimates that customers can begin around 80% through the implementation journey, with much of the remaining work focused on configuration for individual business requirements.

The broader implication is important. The more industry requirements already exist within the platform, the less effort enterprises may need to spend recreating them during implementation.

AI needs context, not just clean data

The industry-specific argument becomes more significant as AI enters ERP environments. AI is showing maturity in areas such as shop floor planning and scheduling, supply chain visibility and dynamic supply chain adjustment. But deploying AI tools without first identifying the business problem risks creating technology without meaningful outcomes.

Infor's Velocity Suite follows a three-stage model: Diagnose, Automate and Optimise. Process intelligence and process mining are first used to identify bottlenecks, constraints and improvement opportunities. Established use cases, including invoice processing, shop floor scheduling and proof-of-delivery automation, can then be applied. Generative AI and ML can subsequently address requirements that are more specific to an individual organisation.

The sequence highlights an important enterprise AI lesson: automation should follow diagnosis, not precede it. There is another requirement - Context.

Clean and consistent data alone may not tell an AI system enough about how a business actually operates. A food and beverage organisation, for example, can spend considerable time consolidating and cleaning information from multiple systems, yet still end up with a generic data repository that lacks meaningful industry context.

That difference becomes critical when AI agents begin making or executing operational decisions. Waste management illustrates the problem. A hospital deals with clinical waste. In food and beverage, waste directly affects production yield. Industrial manufacturing may focus on scrap processing.

An effective AI agent needs to understand these distinctions. For enterprises building AI on top of ERP, the challenge is therefore not simply creating a large pool of clean data. The data must carry enough business context for AI to understand the environment in which it operates and the task it is expected to perform.

More AI agents demand stronger governance

As ERP moves towards agentic AI, governance becomes equally important. One emerging risk is agentic sprawl: organisations deploying growing numbers of AI agents without sufficient oversight. Innovation without governance can increase security and operational exposure rather than accelerate meaningful adoption.

Infor's Agentic Enterprise vision addresses this through a pillar called Governed Velocity. The approach includes explainable AI, AI identity management, identity-enforced MCP boundaries, an agentic rules engine, role-based security and monitoring across transactions and master data.

The principle matters beyond any individual platform. As AI moves from providing information to executing business processes, enterprises need clear controls over what an agent can access, what actions it can perform and when human approval is required.

An AI assistant answering a question carries one level of risk. An autonomous agent acting across business systems carries another. Governance, therefore, cannot be treated as something to add after AI adoption scales. Architectural decisions made early in a modernisation project could determine how safely an organisation can expand autonomous technology later.

From ERP to Enterprise Resource Execution

The longer-term change may be more fundamental. Traditional ERP is evolving towards the concept of Enterprise Resource Execution, or ERX. In this model, ERP remains the secure transactional foundation for financial records, compliance, auditability and core business processes. The greater opportunity emerges from what organisations build above that foundation.

Open APIs connect systems. AI and ML interpret data. Agents orchestrate activities. Analytics becomes more forward-looking and user experiences adapt to what an individual needs at a particular moment. This could also change the role of enterprise users.

Instead of people manually executing every task through predefined forms and workflows, agents could increasingly perform routine work. Humans would remain in the loop as approvers and decision-makers, providing oversight, judgement and accountability.

The transition from ERP to ERX, therefore, is not about abandoning the transactional core. It is about turning that core into a platform from which more work can be intelligently executed.

For CIOs and CTOs, the next ERP transformation may be judged less by whether a new system successfully goes live and more by whether it continuously produces measurable business improvements.

That requires a different order of thinking: define outcomes before selecting capabilities, modernise without outrunning the business, give AI meaningful context and establish governance before autonomy expands.

ERP has traditionally helped enterprises plan and record what the business does. Its next phase could be about helping the business act. That is the real significance of the shift from planning to execution.

Read More:

, discussed how ERP modernisation is shifting towards phased delivery, measurable outcomes and industry-specific technology, while AI is pushing ERP beyond its traditional transactional role towards Enterprise Resource Execution, or ERX.

Modernisation needs measurable outcomes

The big-bang ERP approach is increasingly difficult to justify when businesses cannot wait several years for returns. Instead of beginning with the question of which technology to deploy, organisations need to identify what they want to improve over the next six or 12 months.

That changes the structure of transformation.

Infor's approach, described as value mapping, identifies the financial and operational improvements expected from a project before implementation begins. Business priorities are ranked, less critical requirements can move into later phases and measurable KPIs provide checkpoints along the way.

The underlying principle can be reduced to three steps: identify the outcome, deliver it and measure it. This shifts ERP modernisation from a technology milestone to a business performance exercise. Each phase has to demonstrate value before the organisation moves further along the roadmap.

It also addresses one of the biggest risks in ERP transformation: moving faster than the organisation can handle. Enterprises need to modernise at the speed the business can absorb, rather than simply at the speed technology allows.

For systems sitting at the operational core of an enterprise, this distinction matters. Disruption can affect production, revenue, supply chains and customer delivery. Business continuity, therefore, becomes as much a change management issue as a technical one.

Infor's approach is to build Cloud solutions around specific industries and even micro-vertical requirements. A bakery, for example, does not operate in exactly the same way as a cheese manufacturer, despite both belonging to the food and beverage sector.

This industry-specific foundation can reduce the amount of customisation required during implementation. The company estimates that customers can begin around 80% through the implementation journey, with much of the remaining work focused on configuration for individual business requirements.

The broader implication is important. The more industry requirements already exist within the platform, the less effort enterprises may need to spend recreating them during implementation.

AI needs context, not just clean data

The industry-specific argument becomes more significant as AI enters ERP environments. AI is showing maturity in areas such as shop floor planning and scheduling, supply chain visibility and dynamic supply chain adjustment. But deploying AI tools without first identifying the business problem risks creating technology without meaningful outcomes.

Infor's Velocity Suite follows a three-stage model: Diagnose, Automate and Optimise. Process intelligence and process mining are first used to identify bottlenecks, constraints and improvement opportunities. Established use cases, including invoice processing, shop floor scheduling and proof-of-delivery automation, can then be applied. Generative AI and ML can subsequently address requirements that are more specific to an individual organisation.

The sequence highlights an important enterprise AI lesson: automation should follow diagnosis, not precede it. There is another requirement - Context.

Clean and consistent data alone may not tell an AI system enough about how a business actually operates. A food and beverage organisation, for example, can spend considerable time consolidating and cleaning information from multiple systems, yet still end up with a generic data repository that lacks meaningful industry context.

That difference becomes critical when AI agents begin making or executing operational decisions. Waste management illustrates the problem. A hospital deals with clinical waste. In food and beverage, waste directly affects production yield. Industrial manufacturing may focus on scrap processing.

An effective AI agent needs to understand these distinctions. For enterprises building AI on top of ERP, the challenge is therefore not simply creating a large pool of clean data. The data must carry enough business context for AI to understand the environment in which it operates and the task it is expected to perform.

More AI agents demand stronger governance

As ERP moves towards agentic AI, governance becomes equally important. One emerging risk is agentic sprawl: organisations deploying growing numbers of AI agents without sufficient oversight. Innovation without governance can increase security and operational exposure rather than accelerate meaningful adoption.

Infor's Agentic Enterprise vision addresses this through a pillar called Governed Velocity. The approach includes explainable AI, AI identity management, identity-enforced MCP boundaries, an agentic rules engine, role-based security and monitoring across transactions and master data.

The principle matters beyond any individual platform. As AI moves from providing information to executing business processes, enterprises need clear controls over what an agent can access, what actions it can perform and when human approval is required.

An AI assistant answering a question carries one level of risk. An autonomous agent acting across business systems carries another. Governance, therefore, cannot be treated as something to add after AI adoption scales. Architectural decisions made early in a modernisation project could determine how safely an organisation can expand autonomous technology later.

From ERP to Enterprise Resource Execution

The longer-term change may be more fundamental. Traditional ERP is evolving towards the concept of Enterprise Resource Execution, or ERX. In this model, ERP remains the secure transactional foundation for financial records, compliance, auditability and core business processes. The greater opportunity emerges from what organisations build above that foundation.

Open APIs connect systems. AI and ML interpret data. Agents orchestrate activities. Analytics becomes more forward-looking and user experiences adapt to what an individual needs at a particular moment. This could also change the role of enterprise users.

Instead of people manually executing every task through predefined forms and workflows, agents could increasingly perform routine work. Humans would remain in the loop as approvers and decision-makers, providing oversight, judgement and accountability.

The transition from ERP to ERX, therefore, is not about abandoning the transactional core. It is about turning that core into a platform from which more work can be intelligently executed.

For CIOs and CTOs, the next ERP transformation may be judged less by whether a new system successfully goes live and more by whether it continuously produces measurable business improvements.

That requires a different order of thinking: define outcomes before selecting capabilities, modernise without outrunning the business, give AI meaningful context and establish governance before autonomy expands.

ERP has traditionally helped enterprises plan and record what the business does. Its next phase could be about helping the business act. That is the real significance of the shift from planning to execution.

Read More: