The ERP stories enterprise IT leaders actually read.
Volume 342, Issue 4 · 7 articles from 4 issues
AI Will Transform ERP Into Intelligent Manufacturing Ecosystems: Enlight Metals' Dhananjay Goel
Dhananjay Goel, Managing Director of Enlight Metals, argues that AI agents will revolutionize manufacturing by augmenting traditional ERP platforms rather than replacing them.
Traditional systems excel at data centralization and process standardization but cannot autonomously analyze the massive volumes of data generated by IIoT sensors and connected equipment.
AI agents address this gap by forecasting disruptions, automating workflows, optimizing production schedules, and enabling real-time decision-making across supply chains.
The integration requires clean data infrastructure, workforce training, and robust governance frameworks. As digital twins and edge computing mature, manufacturers adopting AI will gain advantages in efficiency, resilience, and market responsiveness.
Experts outline hidden ERP implementation costs, from extra licensing and integrations to customization and talent gaps.
ERP remains a core enterprise system, but implementations are prone to overruns: Panorama found 30% of projects over budget and Gartner says over 70% miss original goals.
Consultants from Withum, Protiviti, and ElevatIQ describe common hidden costs, including licenses and modules missed in initial bids, selecting a poorly fitting ERP, underestimated integrations with ongoing middleware and API expenses, and costly customizations the buyer must maintain.
Leaders also often overestimate staff expertise and availability, creating talent shortfalls. Thorough upfront planning is presented as the best way to avoid surprise costs.
Clean-core ERP is being reframed as the price of admission to AI, but only about half of organizations are moving.
Years of ERP customization made sense when SAP or Oracle systems would not bend to how a factory ran shifts or a distributor priced products, but it quietly became one of the most expensive decisions companies never explicitly signed off on.
Sunlight Group's Angela Maragkopoulou tells leadership not to sell standardization to the board as technical cleanup but to frame it as the price of admission to AI.
The correction is not yet widespread: IDC research shows the share of organizations preferring a customized approach barely moved, from 52% in late 2024 to 51% by August 2025, which IDC's Mickey North Rizza calls a strategic minority move concentrated in large enterprises feeling technical-debt pain.
COOs face five ERP integration obstacles: data compatibility, cost, resistance, security, and process misalignment.
The article outlines five ERP integration challenges COOs should plan for: incompatible data models across legacy and modern systems; high implementation costs from budget overruns and delays, best managed via disciplined project management and limiting customization; organizational resistance to change requiring sustained change management and phased rollouts; security and compliance risks from increased data flows, needing strong governance and vendor assessments; and process misalignment from departmental silos, solved by mapping and standardizing workflows. Proactive planning turns these obstacles into opportunities.
AI-powered ERP systems let employees use natural language instead of navigating menus to get work done.
The article explains that traditional ERP systems require users to navigate multiple menus and rigid workflows for routine tasks, consuming significant time.
AI-powered ERP platforms, illustrated by ZYNO, let employees request information conversationally, auto-generate customized reports, and automate repetitive approvals across departments.
Key features include a natural-language smart assistant, prompt-based reporting, workflow automation, and cross-module intelligence linking finance, procurement, HR, and inventory data.
This shifts interaction from "search, navigate, filter" to "ask, understand, act," turning ERP from a transaction-recording system into an intelligent platform that cuts manual work and speeds decisions.
CIOs must plan beyond ERP go-live to avoid stalled momentum, urging capability owners and workforce shifts.
The article warns of an "ERP hangover" where momentum stalls after go-live once the implementation roadmap disappears.
It recommends reorienting operating models around capability managers who own continuous improvement, shifting from project execution to ongoing product optimization.
It also urges rethinking workforce strategy as automation changes which roles are needed, emphasizing orchestration over transactional work.
Finally, it advises organizations to formally distinguish project work from product work and define transition criteria. CIOs should name capability owners and launch roadmaps before hypercare ends.
Five questions executives should ask ERP vendors about AI agents covering value, audit trails, security, and scale.
The article lists five questions C-suite leaders should ask ERP vendors before deploying AI agents: - what tasks the agent performs and its real business value; - whether comprehensive audit trails exist for inputs, reasoning, and actions to support compliance and bias detection; - whether data security meets regulations like GDPR and HIPAA without training external models on corporate data; - whether the agent has proven enterprise-scale deployments over 3-6 months; and how well it integrates with existing workflows and systems.
James stresses keeping high-stakes decisions under human control.