Connect JD Edwards to AI with the JDE MCP Server
Give ChatGPT, Claude, Copilot, and AI agents secure, governed access to JD Edwards data, applications, orchestrations, and business processes.
Ask questions. Find information. Execute approved actions. Automate work across EnterpriseOne.
Enterprise AI Platforms
Book a Demo
See the JDE MCP Server in action.
"*" indicates required fields
What is MCP?
A Standard Way to Connect AI to the Systems That Run Your Business
Model Context Protocol (MCP) is an open standard that gives AI applications a consistent way to connect with external systems, data, tools, and workflows.
For JD Edwards organizations, MCP provides a standardized way for approved AI applications and agents to interact with EnterpriseOne — from retrieving business information to executing approved processes. Think of it as the connection between the AI your organization wants to use and the JD Edwards systems that run your business.
AI Applications & Agents
🔌 JDE MCP Server
Access • Tools • Governance
↕
JD Edwards EnterpriseOne
Data • Applications • Orchestrations • Business Processes
AI Can Answer Questions. The Bigger Opportunity Is Giving It Access to the Systems That Run Your Business.
For JD Edwards organizations, that raises important questions: How does AI access the right JDE information? How can it take action without bypassing established business processes? How do you control which tools each user or AI agent can access? How do you avoid building a new integration for every AI use case?
The JDE MCP Server creates a governed connection between AI and EnterpriseOne so organizations can begin putting AI to work across existing JD Edwards processes.
📈
Extend Your JDE Investment
⚡
Accelerate AI Initiatives
Expose reusable JDE capabilities as AI-ready tools instead of approaching every use case as a separate integration project.
🛡
Maintain Control
🔮
Keep Your AI Strategy Flexible
Create a standards-based connection to JD Edwards that can support different AI applications, agents, and frameworks. Not locked into one AI vendor.
See What Becomes Possible
The Fastest Way to Understand MCP Is to See What Users Can Do with It
Click any module below to see how a natural language request becomes a JDE action.
- ChatGPT • Circular Edge JDE MCP Server
You
Create a PO for supplier ABC for 100 units of Item 123.
JDE MCP Server → Orchestration: Create Purchase Order
Purchase order “OP-2026-08491” created successfully.
Supplier: ABC Materials Inc (Vendor #48201)
Item: 123 · Qty: 100 units
Unit Cost: $12.50 · Total: $1,250.00
Status: Pending approval — notification sent to approver
Expected Delivery: 14 days from approval
Behind each request: AI identifies an approved tool → JDE MCP Server connects to EnterpriseOne → JD Edwards executes the inquiry or business process → Result returned to user
- Claude • Circular Edge JDE MCP Server
You
What’s the inventory level for Item 123 across all warehouses?
Item 123 — Current On-Hand by Branch Plant
🏭 Branch Plant 01 (Somerset NJ): 847 units — ✓ Above safety stock
🏭 Branch Plant 02 (Houston TX): 312 units — ✓ Above safety stock
🏭 Branch Plant 03 (Chicago IL): 89 units — ⚠ BELOW safety stock (min: 200)
Total on-hand: 1,248 units • Total available: 1,104 units
Recommend replenishment transfer or PO expedite for Chicago.
Behind each request: AI identifies an approved tool → JDE MCP Server connects to EnterpriseOne → JD Edwards executes the inquiry or business process → Result returned to user
- Microsoft Copilot • Circular Edge JDE MCP Server
You
Show me the open sales orders for Customer ABC.
Customer ABC — Open Sales Orders (3 found)
- SO-2026-14201 · 500 units Item 123 · Status: In Picking · Ship by Aug 14
- SO-2026-14388 · 200 units Item 456 · Status: Backordered · ETA Aug 21
- SO-2026-14521 · 1,000 units Item 789 · Status: Entered · Ship by Aug 28
Total open value: $48,750. Would you like shipment details or to escalate the backorder?
Behind each request: AI identifies an approved tool → JDE MCP Server connects to EnterpriseOne → JD Edwards executes the inquiry or business process → Result returned to user
- ChatGPT • Circular Edge JDE MCP Server
You
Show unpaid invoices over $50,000.
4 unpaid invoices over $50,000 — Total: $428,750
- INV-2026-00841 · Acme Corp · $182,500 · 48 days past due
- INV-2026-00729 · Global Distributors · $97,200 · 41 days past due
- INV-2026-00814 · Pacific Supply Co · $89,050 · 35 days past due
- INV-2026-00788 · MidWest Trading LLC · $60,000 · 31 days past due
Would you like payment reminders sent or these escalated to collections?
Behind each request: AI identifies an approved tool → JDE MCP Server connects to EnterpriseOne → JD Edwards executes the inquiry or business process → Result returned to user
- Azure AI Foundry • Circular Edge JDE MCP Server
You
What’s the status of Work Order 12345?
Work Order 12345 — Status Summary
Type: Manufacturing Work Order • Item: 123 — Assembly A
Ordered Qty: 500 • Completed: 312 (62%) • Scrapped: 8
Status: In Process • Current Operation: Op 30 — Final Assembly
Target Completion: August 15, 2026
Material Shortages: Item 9102 — 13 units short. PO expedite recommended.
Behind each request: AI identifies an approved tool → JDE MCP Server connects to EnterpriseOne → JD Edwards executes the inquiry or business process → Result returned to user
Procurement
“Create a PO for supplier ABC for 100 units of Item 123.”
Inventory
“What’s the inventory level for Item 123 across all warehouses?”
Sales
Finance
Manufacturing
Built for JD Edwards
Every EnterpriseOne Capability AI Needs — All Supported
The JDE MCP Server is designed to expose the EnterpriseOne capabilities AI applications and agents need to retrieve information and execute approved business processes. Use the JDE services and processes you already have while creating new ways for employees and AI agents to interact with EnterpriseOne.
| JD Edwards Capability | Supported | Read | Write / Execute | Notes |
|---|---|---|---|---|
| Orchestrations Most Popular JDE Orchestrator workflows and multi-step business processes | ✓ | ✓ | ✓ | AI invokes orchestrations by name with natural language parameters. |
| AIS APIs JDE Application Interface Services RESTful endpoints | ✓ | ✓ | ✓ | Full AIS API surface across all JDE modules. Authentication handled server-side. |
| Form Service Requests (FSR) JDE interactive application forms via AIS | ✓ | ✓ | ✓ | AI interacts with JDE forms exactly as a human user would. |
| Data Service Requests (DSR) Direct JDE business data queries via AIS | ✓ | ✓ | ✓ | Natural language filters translated into JDE data queries automatically. |
| Business Functions JDE server-side business function calls (BSVVs) | ✓ | ✓ | ✓ | Execute JDE business functions including calculations and validations. |
| Watchlists JDE operational intelligence dashboards | ✓ | ✓ | — | AI surfaces watchlist exceptions and operational metrics in conversation. |
| Custom Objects JDE user-defined custom business objects | ✓ | ✓ | ✓ | Your custom JDE objects accessible to AI exactly like standard objects. |
| Notifications JDE workflow notifications and approvals | ✓ | ✓ | ✓ | AI reads and acts on pending approvals and workflow notifications. |
| CafeOne Integrations JDE CafeOne embedded application links | ✓ | ✓ | — | CafeOne context available for enriched JDE data access in AI workflows. |
| User Defined Objects (UDOs) Custom saved queries, reports, notifications | ✓ | ✓ | ✓ | AI discovers and executes JDE UDOs, making your customizations AI-readable. |
Give AI the Right Tools
Access to JD Edwards does not have to mean access to everything. The JDE MCP Server can make specific tools available based on the user, role, application, or AI agent. The result is an AI experience designed around the same business responsibilities and controls your organization already expects from its enterprise systems.
🏭 Warehouse Manager
- Inventory Inquiry
- Transfer Inventory
- Cycle Count
- Inventory Receipt
- Create Supplier
- Modify Financial Records
💬 Customer Service
- Customer Inquiry
- Order Inquiry
- Shipment Status
- Invoice Inquiry
- Inventory Adjustments
- Supplier Maintenance
Access is designed around the responsibilities and controls your organization already expects from its enterprise systems. Tool permissions are configured per user, role, or AI agent and apply to every request — human or automated.
Give AI Access Without Giving Up Control
Connecting AI to ERP requires more than making an API available. Organizations need visibility and control over who can access JD Edwards, which tools they can use, and what actions can be executed. The JDE MCP Server supports enterprise controls including:
Existing JDE Security
Work with the access and roles already established within EnterpriseOne. No need to rebuild security from scratch.
Role-Based Access
Make tools available according to the responsibilities of individual users and roles in your organization.
Tool-Level Permissions
Control which JDE capabilities each user or AI agent can discover and invoke. Granular control at the tool level.
Read vs. Write Controls
Separate information retrieval from tools that create or modify transactions in JD Edwards.
Human Oversight
Require approval or confirmation for processes that should not execute autonomously. Humans stay in the loop where required.
Audit Logging
Maintain visibility into AI activity and the tools and actions used within JD Edwards. Every request is logged with full context.
Credential Protection
Keep JD Edwards authentication information protected from the AI model at all times. Credentials never leave the server layer.
Encrypted Communication
Protect information as requests move between AI applications, the MCP layer, and EnterpriseOne at every step.
Five Steps from Natural Language Request to JDE Result
The JDE MCP Server sits between approved AI applications and JD Edwards, translating AI requests into governed access to EnterpriseOne tools and processes.
1
Ask
“Show inventory for Item 123 across all branch plants.”
User or AI Agent Makes a Request
Any natural language request from a user or an AI agent triggering a workflow.
Discover
AI Identifies the Approved JDE Tool
The AI identifies the appropriate approved tool that can fulfil the request.
3
Govern
MCP Server Applies Access and Policies
The JDE MCP Server applies the relevant access, permissions, and policies before JDE is reached.
Execute
JDE Performs the Inquiry or Transaction
JD Edwards performs the inquiry, orchestration, or transaction using the appropriate EnterpriseOne service.
Respond
Result Returned to User or Agent
The result is returned to the user or agent through the AI experience in plain language.
JDE MCP Server Architecture
🌐 Enterprise AI Platforms
↕
🔌 JDE MCP Server
Tool Discovery • Access Control • Governance • Audit
↕
🏭 JD Edwards EnterpriseOne
AIS APIs
Already Using Orchestrator? You May Already Have Valuable Building Blocks for AI.
An existing orchestration might already do exactly what an AI agent needs to do. With the JDE MCP Server, that established process can become an approved tool that an AI application or agent can discover and invoke.
A user could simply ask:
Get more value from the orchestrations you have already built while giving AI a controlled, governed way to interact with established JDE processes.
The JDE MCP Server Can Support Everything from Simple Conversational Access to Advanced AI-Driven Workflows
A Reusable AI-Ready Layer Over Approved EnterpriseOne Capabilities
Connecting AI to JD Edwards through individual APIs can require separate development, integration logic, and governance for each AI application or use case. The JDE MCP Server creates a reusable AI-ready layer over approved EnterpriseOne capabilities.
🔌
AI-Ready JDE Tools
Make JDE capabilities understandable and discoverable to AI applications and agents. AI can find and use the right tool without custom development for each use case.
🛡
Centralized Governance
🔁
Reusable Connections
🏭
Built Around EnterpriseOne
📈
Designed for Expansion
Start with a focused use case and add additional tools and workflows as your AI strategy develops. The architecture scales with your ambitions.
You Do Not Need to Connect Every JD Edwards Process to AI at Once
Make availability and item information easier to access. Reduce time spent navigating JDE to answer simple inventory questions.
Give teams faster access to orders, shipments, and fulfillment status. Reduce customer service response time with conversational JDE access.
Connect AI with approved purchasing processes and orchestrations. Reduce manual PO creation and status lookup effort.
Make AR, AP, and invoice information easier to research. Give finance teams conversational access to JDE without complex screen navigation.
Give administrators AI-assisted access to approved diagnostic and operational tools. Reduce time on routine JDE monitoring and lookups.
Start with one process. Prove the value. Expand from there. Circular Edge will help you identify the right starting point for your JDE environment.
Book a Demo →Start with a defined process. Determine what AI needs access to, how JD Edwards should execute the request, and what controls need to be in place. Circular Edge helps organizations answer those questions and move from use case to working solution.
JDE MCP — Common Questions
What is a JDE MCP Server?
A JDE MCP Server is a governed connection layer that gives AI applications and agents secure access to JD Edwards EnterpriseOne. It uses the Model Context Protocol (MCP), an open standard that gives AI a consistent way to connect with external systems. For JD Edwards organizations, the JDE MCP Server translates AI requests into governed access to EnterpriseOne tools and processes, covering orchestrations, AIS APIs, form service requests, data service requests, business functions, and custom objects.
What is MCP and why does it matter for JD Edwards?
Model Context Protocol (MCP) is an open standard that gives AI applications a consistent way to connect with external systems, data, tools, and workflows. For JD Edwards organizations, MCP provides a standardized way for approved AI applications and agents to interact with EnterpriseOne without building a new integration for every AI use case. It removes the need for custom API development per AI platform and per JDE process.
Which AI platforms work with the JDE MCP Server?
The Circular Edge JDE MCP Server works with ChatGPT, Claude, Microsoft Copilot, Azure AI Foundry, Google Gemini, custom AI agents, and enterprise AI platforms that support MCP. It creates a standards-based connection to JD Edwards that can support different AI applications, agents, and frameworks without requiring a separate integration for each one.
How does the JDE MCP Server handle security and governance?
The JDE MCP Server supports enterprise controls including existing JDE security, role-based access, tool-level permissions, read vs. write controls, human oversight requirements, audit logging, credential protection, and encrypted communication. JD Edwards credentials are never shared with AI models. Organizations control what AI can access, which actions it can perform, and where approvals or additional controls are required.
What JD Edwards capabilities can AI access through the JDE MCP Server?
The JDE MCP Server supports JDE Orchestrations, AIS APIs, Form Service Requests, Data Service Requests, Business Functions, Watchlists, Custom Objects, Notifications, CafeOne Integrations, and User Defined Objects (UDOs). Organizations can use the JDE services and processes they already have while creating new ways for employees and AI agents to interact with EnterpriseOne.
Where should an organization start with JDE MCP?
Start with a business process where employees spend significant time searching for information, navigating applications, or completing repetitive work. Common starting points include inventory availability inquiries, order management and shipment status, procurement workflows, AR and AP research, and JDE system administration. Start with one process, prove the value, and expand from there. Circular Edge can help identify the right starting point for your environment.