Your JD Edwards, ready to talk to AI.
XP Flow MCP connects JD Edwards EnterpriseOne to AI assistants such as ChatGPT, Claude, Microsoft Copilot and Gemini. Ask a general or specific question about your JDE data, let AI analyse it, and even trigger orchestrations to run entire processes or perform specific tasks. All in plain language, in dialogue with your AI client — without building a new integration for every AI tool your organisation uses.
What is MCP?
One connection. A new way to interact with JD Edwards.
MCP — the Model Context Protocol — is an open standard that enables AI applications to connect to external systems, tools and data through a common interface.
Instead of building and maintaining a separate AI integration for every platform, MCP provides a common way for compatible AI applications to discover and use the capabilities you choose to expose.
XP Flow MCP brings that model to JD Edwards EnterpriseOne.
It acts as an MCP server — a bridge between JD Edwards and compatible AI applications — giving them controlled access to JDE Orchestrations, data and attachments.
And because the connection is based on an open protocol, your JD Edwards AI strategy does not have to depend on a single AI vendor.
What can XP Flow MCP do?
Run your existing orchestrations
Every orchestration you choose to publish from your JDE environment can become a capability an AI assistant can call directly.
XP Flow MCP discovers available orchestrations from your environment, rather than requiring each one to be manually recreated as a separate AI integration.
Your existing JDE business logic remains where it belongs: in JD Edwards.
Query JD Edwards data
Ask about the status of a specific customer, item or order and let AI assess the situation and provide recommendations. Ask complex questions, and AI will search across your data, link results together, and compile them into text, documents or visuals — all tailored to your exact needs in the moment. This gives AI access to the relevant information without requiring a separate BI report, analysis or consolidation for every question. This is true ad-hoc reporting.
How it works
Your AI assistant asks. XP Flow MCP connects. JD Edwards does the work.
XP Flow MCP sits between your AI assistant and JD Edwards EnterpriseOne, giving the AI a controlled way to use the JDE capabilities you choose to make available.
1. Ask in your AI assistant
A user asks a question or requests an action in ChatGPT, Claude, Microsoft Copilot, Gemini or another compatible AI application.
2. XP Flow MCP connects the request to JD Edwards
XP Flow MCP makes the relevant JDE capabilities available through MCP and routes the request to the appropriate function.
3. JD Edwards provides the data or performs the action
Depending on the request, JD Edwards can:
-
Run an orchestration — using your existing JDE business logic and processes.
-
Query data — retrieving read-only information from JDE tables and business views.
-
Retrieve attachments — accessing documents, notes and images through JDE's attachment services.
The result is returned to the AI assistant and presented to the user in plain language.
AI assistant → XP Flow MCP → JD Edwards EnterpriseOne
Secure
Your AI can only access the data and orchestrations you grant it, controlled either through JDE's standard role- and user-level security or directly in the MCP configuration. Since the MCP server can be hosted within your own environment, no data leaves your organisation.
Scalable
Guided by your instructions, and over time, your AI will get better at interpreting your questions, your data, and which actions to recommend or carry out. If AI lacks the tools to correct, optimise or handle a situation, you can easily extend its capabilities by creating new orchestrations — AI will discover them automatically and, without further instruction, be able to run them.
From using JDE to simply asking JDE
What if working with JD Edwards could start with a question?
Instead of navigating screens, running reports or building separate integrations, users can interact with the JDE capabilities and information you choose to make available — directly from the AI tools they already use.
For example:
"What's the purchasing status of item 210?"
XP Flow MCP searches purchase orders for the item and returns the status, promised date and quantity received. Ask a follow-up about how the current situation affects open sales orders and forecast. Resolve it by asking AI to place a purchase order.
"Create a chart and a pivot table of the 10 items with the best margin last quarter."
AI compiles the data, presents it directly in the response, and offers recommendations.
"Which items in branch/plant M30 are below their reorder point?"
AI runs a filtered, read-only query against the relevant item and branch/plant data, suggests actions, and can place the necessary purchase orders.
"Which users have used which programs over the past year?"
AI compiles JDE usage data. Give AI access to your licence agreement and get an immediate analysis of whether you're correctly licensed.
You ask. JD Edwards answers.
One MCP connection. Multiple AI experiences.
AI platforms are changing quickly. Your ERP architecture shouldn't have to change every time they do.
XP Flow MCP provides a common layer between JD Edwards and MCP-compatible AI applications.
That means the capabilities you expose from JD Edwards can be made available to different AI experiences without designing a completely new JDE integration for each one.
Today that might mean ChatGPT, Claude, Microsoft Copilot or Gemini. Tomorrow it may be something else.
The important part is that JD Edwards remains at the centre — while MCP provides a standard way for AI to interact with it.
Control what AI can see — and what it can do.
Connecting AI to an ERP requires more than connectivity.
It requires control.
XP Flow MCP is designed around the capabilities and information you deliberately make available from JD Edwards. It can be hosted against a scoped JDE service account, limited to the orchestrations and data you are comfortable exposing.
Direct queries against JD Edwards data are read-only, while actions can be handled through the JDE orchestrations you choose to make available.
The AI doesn't replace JD Edwards business logic. It provides a new way to interact with it.
This distinction matters: the goal isn't to give an AI unrestricted access to your ERP. It's to give users a simpler interface to the specific JDE capabilities you decide should be available.
Built for JD Edwards. By people who know JD Edwards.
Connecting an AI assistant to an ERP isn't primarily an AI problem.
It's an ERP problem too.
You need to understand what users should be allowed to see, what they should be allowed to do, where the business logic belongs — and how JD Edwards actually works.
xperitus has worked with JD Edwards organisations for more than 25 years and has been an Oracle Partner since 2008.
Today, our team of 35+ ERP consultants works with manufacturing, distribution and project-based organisations across the Nordics and beyond.
XP Flow MCP grew out of a simple question we increasingly heard from JD Edwards customers:
“Can we just ask it?”
We thought the answer should be yes.
So we built the bridge between the ERP we know and the AI tools our customers are starting to use.
Ready to have a conversation with JD Edwards?
You don't need to start with a large AI project.
Start with one useful question, one existing orchestration or one JDE process your users deal with every day.
We'll show you how it can work through XP Flow MCP.
Talk to us about connecting your JD Edwards environment to AI.
MCP stands for Model Context Protocol. It is an open protocol designed to provide a standard way for AI applications to connect to external tools, systems and data sources.
For JD Edwards organisations, this means you can create a common connection between EnterpriseOne and compatible AI applications instead of designing a completely separate integration for every AI platform.
XP Flow MCP is xperitus's MCP server for JD Edwards EnterpriseOne. It makes selected JDE capabilities available to compatible AI applications through the Model Context Protocol.
Depending on your configuration, this can include JDE orchestrations, read-only queries against tables and business views, and access to attachments.
XP Flow MCP is built around the Model Context Protocol rather than one specific AI vendor. It can therefore be used with AI applications that support the protocol, including MCP-enabled experiences involving platforms such as ChatGPT, Claude, Microsoft Copilot and Gemini.
Exact functionality depends on the MCP support and configuration of the AI application being used.
No. It complements it.
XP Flow MCP is designed to work with orchestrations already published in your JD Edwards environment, making them available for compatible AI applications to call through MCP.
Your existing orchestrations remain in JD Edwards rather than being recreated as separate AI integrations.
Think of Orchestrator as defining what JDE can do, while MCP provides a standard way for AI to ask it to do it.
XP Flow MCP can expose JDE orchestrations that perform actions in EnterpriseOne. Direct queries against JD Edwards tables and business views are read-only.
This makes it possible to distinguish between asking JD Edwards for information and asking it to run an orchestration.
What an AI application can ultimately access or execute depends on the orchestrations and data made available through your XP Flow MCP setup.
No. The purpose of XP Flow MCP is not to expose an entire ERP environment to an AI.
XP Flow MCP can be hosted against a scoped JDE service account, limited to the orchestrations and data you are comfortable exposing.
This means you can start with a small, clearly defined scope and expand it as your use of XP Flow MCP grows.
Yes. XP Flow MCP can make JD Edwards attachments available through JDE's attachment services, including documents and images where supported.
This means a user can ask not only about a transaction but also about supporting information associated with it.
No. XP Flow MCP is designed to work with orchestrations already published in your JD Edwards environment.
That means existing investments in JD Edwards Orchestrator can become part of your AI strategy rather than being replaced by it.
That is exactly the problem MCP is intended to reduce.
Instead of creating a unique JDE integration for every AI application, XP Flow MCP provides a common MCP interface to the capabilities you expose from EnterpriseOne.
Support and configuration can differ between AI platforms, but the JD Edwards side does not have to be redesigned from scratch each time.
Start small.
Choose a JDE use case that is easy to understand and useful to your organisation – for example checking an order status, finding customer information or calling an existing orchestration.
We start with one team and one clear use case, with XP Flow MCP hosted against a scoped JDE service account limited to the orchestrations and data you are comfortable exposing.
Once it has earned its place, the setup can be widened to more use cases, more teams and eventually per-user access as your rollout grows.