Infor is making the case that manufacturers need to address their underlying ERP architecture before they can get meaningful value from embedded AI agents.
In a four-part series aimed at CFOs, CIOs, COOs, and CEOs in metal fabrication, the company connects Infor CloudSuite Industrial Enterprise with Infor Velocity Suite and its Industry AI Agents. Across the posts, the sequence is consistent: establish connected operational data, identify process problems, and then introduce automation and agents into those workflows.
That puts the ERP foundation ahead of the AI. Infor presents industry-specific processes and connected data as prerequisites for expanding automation across finance and manufacturing.
Why Infor Puts the ERP Foundation Before AI
Infor identifies fragmented systems and accumulated customization as barriers to modernization and AI adoption.
Its CIO-focused guidance describes manufacturers running ERP alongside manufacturing execution, product lifecycle management, engineering, quoting, and supplier systems connected through years of custom integrations. Moving that environment to the cloud without changing the underlying architecture can preserve much of the same technical debt.
Infor positions CloudSuite Industrial Enterprise around requirements specific to industrial manufacturing, including dimensional inventory, project costing, material pricing, configure-price-quote processes, and connections between ERP, PLM, and manufacturing execution. The company says providing those functions within the platform can reduce the customization that accumulates around more generic ERP deployments.
That foundation also affects AI. Agents working across finance or production depend on consistent information about orders, inventory, projects, engineering changes, suppliers, and costs. Fragmented data that previously created reporting and integration problems can also limit the context available to automated systems.
Infor’s CEO-focused article makes a similar connection between operational performance and technology architecture, positioning connected, trusted data as the basis for applying AI to business processes.
Process Mining Comes Before the AI Agents
Infor Velocity Suite starts by examining how work currently happens before introducing automation. The suite follows what Infor describes as a discover, build, measure, and expand approach. Process Mining identifies bottlenecks and inefficient workflows first, creating a basis for deciding where automation could deliver value.
In finance, Infor’s CFO guidance highlights agents covering project accounting, accounts payable, general ledger, accounts receivable, and other financial processes. On the operational side, agents can address production orders, inventory, quality, purchasing, and supplier exceptions.
Infor’s Agentic Orchestrator coordinates agents operating across those processes, while the company says automated actions remain traceable for governance and oversight.
The sequence is important because it makes process selection part of the AI deployment model. Instead of starting with an agent and looking for tasks to automate, Infor starts with the workflow and identifies where automation could change its performance.
Metal Fabrication Gives Infor an Industry-Specific AI Case
Metal fabrication provides Infor with a practical example of why the underlying ERP data matters.
Raw materials can represent around 60% of total cost in the sector, according to Infor. Its material-pricing functionality tracks base and actual prices inside purchasing and sales documents and can apply surcharges when input costs change. Standard and actual costing capture materials, labor, and overhead, while project costing gives finance teams visibility into cost-to-complete before a job closes.
Operational requirements are similarly specific. Dimensional inventory can track characteristics such as length, width, and coil diameter, while bill-of-material and routing structures account for yield and scrap.
Infor’s COO-focused guidance also connects ERP with manufacturing execution and engineering systems so production teams can work with current job, quality, material, and product information.
Those structures give agents more specific context. Production agents need reliable scheduling and job data, inventory agents need material availability, and quality agents need access to production and non-conformance information.
Infor’s AI proposition therefore depends as much on the architecture underneath the agents as on the agents themselves. Technical debt that previously increased maintenance and upgrade costs can become an AI constraint when automation has to operate across the same fragmented processes.
What This Means for ERP Insiders
AI changes the cost of technical debt. Integrations that were once mainly a maintenance burden can become barriers to scaling automation, giving companies another reason to address architecture during ERP modernization.
Process selection becomes a governance decision. Mapping workflows before deploying agents creates a defensible basis for deciding where autonomy belongs and gives business and technology teams a shared framework for expansion.
Industry fit matters more as autonomy increases. As software begins acting on operational data, differences in how accurately an ERP represents sector-specific objects and processes can have consequences beyond user convenience.
