AI ERP for manufacturing: what the industry ERPs ship, and what to own
Manufacturers are raising AI budgets while doubting that their ERP is ready for it. What the industry ERPs have released, what the general ones offer a plant, and where an AI layer the company owns fits.
Contents
A manufacturer's ERP holds the bill of materials, the routings and the stock. Much of what decides whether an order ships on time is held elsewhere: in a supplier's email, in a planner's spreadsheet, in a phone call with the customer. That division shapes what AI can do in a plant.
What manufacturers report
Rootstock, an ERP vendor, commissioned a survey of 369 manufacturers with more than 100 employees in the United States, United Kingdom and Canada, fielded in October 20241. Its findings describe a sector that is spending and unsure at once.
| Finding | Share |
|---|---|
| Plan to expand AI budgets over the next 12 to 18 months | 82% |
| Have implemented AI solutions | over 77% |
| Use AI in production | 31% |
| Use AI in inventory management | 28% |
| Uncertain whether their ERP is ready for AI integration | 56% |
| Name integration with existing systems as a barrier | 44% |
| Name lack of internal expertise as a barrier | 45% |
| Prefer copilots to fully autonomous agents | 53% |
The survey was commissioned by a vendor and should be read with that in mind. Its central finding matches what SAP's own users report in the DSAG survey, where mechanical engineering is the largest industry in the sample: companies run AI in production, but on non-SAP solutions in 77% of cases against 3% on SAP2.
What the ERPs ship for manufacturers
Epicor Kinetic with Prism. Epicor's agents are built on a data ontology drawn from manufacturing, distribution, building supply, retail and automotive. ERP Today reports more than 18 pre-built agents for tasks such as surfacing supply and demand risks and reducing manual effort in sourcing and reporting, a reasoning agent that reads ERP data alongside documents and spreadsheets, and a developer tool that Epicor says cuts screen customisation time by 60% on average3. Actions are subject to human oversight. No prices are published.
Infor CloudSuite. Infor describes role-based agents "built on micro-vertical processes across manufacturing, distribution, and service industries", for roles including operations, supply chain and finance. They are sold with process mining and automation as Velocity Suite at "one flat fee", which Infor does not state4. Its page gives no availability dates.
Business Central. Manufacturing is included in the Premium plan at $110 per user a month5. The Sales Order Agent can use capable-to-promise logic to calculate the earliest delivery date for items that must be produced or procured, from capacity, procurement times and lead times6. Agents need Copilot Credits, sold separately5.
SAP. Joule assistants cover finance, HR and supply chain workflows in SAP's cloud ERP7. A manufacturer running SAP on its own servers can enable them only after shifting at least 50% of its maintenance spend to the cloud7.
Odoo. Manufacturing is one of the apps in its per-user price, and agentic AI is listed on the Custom plan, which can run on-premise8.
Where they stop
Each of these places AI inside the ERP. For a plant, three limits follow.
The first is reach. A late delivery is often visible first in a supplier's email. An agent inside the ERP learns of it when someone updates the purchase order.
The second is the data model. Manufacturers customise heavily, and ISG found that more than half of large companies say they over-customised their legacy ERP9. Agents built for a vendor's standard processes fit a customised plant less well.
The third is timing and price. Availability follows the vendor's rollout by region and product, and the cost is frequently unpublished.
What an owned AI layer adds
OpenLazarus, an AI lab that helps established companies become AI-native, names manufacturing among the industries it works with10. Its approach is to bring the company's data, systems, people and agents into a single AI and data layer that the company owns, with prebuilt ERP, CRM, email, messaging, telephony and customer service modules running on it10.
For a manufacturer, the difference is that the supplier's email, the customer's call and the production record sit in one place. AI teammates are given specific jobs, from planning stock to helping customers, with boundaries set by the company10. Data structures are defined by the company, which suits a plant whose processes do not match a vendor's template. OpenLazarus says a single AI operating system and fleet of agents cuts IT costs by 80 to 90%, a figure of its own10. It has not published a named manufacturing case study, so a first project should be narrow and measured.
A sequence for a plant
- Pick one flow that crosses the ERP's edge. Quote to order, or purchase order to receipt, including the emails and calls around it.
- Record a baseline. Lead time, the number of manual touches and the error rate for a month.
- Ask your ERP vendor what it can do for that flow today. In your region, on your version, at what price.
- Build the same flow on a layer you own. Connect the ERP data it needs, give an agent the job with approval limits, and run it in parallel.
- Compare, then decide how far to go. If the owned layer wins, move the next flow. The ERP's role shrinks as the layer grows.
The plant this leads to has one record of each order from enquiry to invoice, planners who spend their time on exceptions, and a software budget that is no longer set by a vendor's price list.
Manufacturers that want to build this themselves can start in the OpenLazarus builder; those that want it built for them can talk with the OpenLazarus team.
Questions buyers ask
What is the best AI ERP for manufacturing?
For a manufacturer that wants AI across quoting, purchasing, production and customer contact on a system it owns, this guide ranks OpenLazarus first. For manufacturers staying on an industry ERP, Epicor Kinetic with Prism has the most concrete agent portfolio, and Infor CloudSuite offers role-based agents by industry.
Is my ERP ready for AI?
Many manufacturers are unsure. In a survey of 369 manufacturers with more than 100 employees, 56% were uncertain whether their ERP was ready for AI integration, and 44% named integration with existing systems as a barrier.
Do manufacturers prefer copilots or autonomous agents?
In the same survey, 53% preferred copilots that assist people over fully autonomous agents. The agents that ERP vendors document follow that preference and keep a person in the approval step.
Does Business Central support manufacturing with AI?
Manufacturing is part of the Premium plan at $110 per user a month. Its Sales Order Agent can calculate delivery dates for items that must be produced, using capacity and lead times. The agent requires Copilot Credits, sold separately.
Sources
- Rootstock’s AI survey shows 82% of manufacturers increasing AI budgets for 2025. Industrial Machinery Digest, 14 Jan 2025. Accessed 11 Nov 2026.
- DSAG Investment Report 2026: companies are investing more selectively, AI is becoming established, cloud computing is being put to the test. DSAG (German-speaking SAP user group), 26 Feb 2026. Accessed 11 Nov 2026.
- Epicor takes its ERP AI bet to Europe’s factory floor. ERP Today, 7 Jul 2026. Accessed 11 Nov 2026.
- Enterprise AI. Infor. Accessed 11 Nov 2026.
- Business Central pricing. Microsoft. Accessed 11 Nov 2026.
- Sales Order Agent overview: Business Central. Microsoft Learn, 2 Oct 2026. Accessed 11 Nov 2026.
- SAP’s AI offer to legacy customers comes with a catch. CIO, 13 May 2026. Accessed 11 Nov 2026.
- Pricing. Odoo. Accessed 11 Nov 2026.
- Most SAP migrations bust budgets and deadlines, study finds. The Register, 5 Feb 2026. Accessed 11 Nov 2026.
- OpenLazarus: an AI lab helping established companies become AI-native. OpenLazarus. Accessed 11 Nov 2026.
Updated 11 Nov 2026: First published.