Six ways to put AI to work, for the people who use apps and the people who build them. Your users get AI in the tools they already know, and JD Edwards stays isolated from AI.
AI runs where your people already work: in their AI assistant, inside EASYProcess apps, as agents that get work done, inside everyday processes, and inside the EASYProcess IDE.
Use your EASYProcess apps from ChatGPT, Claude and more.
Learn more →A chat panel in every EASYProcess app, powered by the AI you choose.
Learn more →Agents work toward a goal, using EASYProcess Functions as tools.
Learn more →AI built into everyday processes, like AP invoice processing.
Learn more →In EASYProcess, business logic lives in Functions. The EASYProcess MCP server turns the Functions you choose into tools that ChatGPT, Claude and other AI assistants can use. Users work with EASYProcess apps from the AI console they already know, with no need to log in to EASYProcess apps.
Each Function is defined once, with inputs, outputs and a description. Each one is a REST API with a clear contract.
POST../Function/POInquiry
One API for pages, chat and agents. The Open PO Inquiry page and AI both call POInquiry: change it once, and both get the change.
Each exposed Function is a tool, and its description guides the AI. Users stay in their AI console.
POST../Function/NotifySupplier
Opening business apps to AI raises fair questions from IT. EASYProcess™ answers them by design.
Expose only the Functions you select. Everything else stays invisible to AI.
The description you write for each Function is what AI reads to decide when and how to use it.
Sign in once with your existing login: Microsoft 365, Okta or your EASYProcess account. EASYProcess is then there in every AI conversation.
Every request carries the user's identity. AI sees only what that user is allowed to see.
AI talks only to EASYProcess, never directly to JD Edwards or your databases. EASYProcess stands between AI and your enterprise data.
A customer is on the line. The agent asks their AI assistant for the order's status and tracking number, and the answer comes from EASYProcess in the same conversation.
Illustrative example
Illustrative example
One click on the AI Chat button in the app header opens a chat panel on the left side of the screen. The chat calls the same Functions as the app.
An AI agent works toward a goal. It decides the steps, uses EASYProcess Functions as tools and checks the results until the job is done.
You describe the outcome in plain language, not the steps.
It acts through EASYProcess Functions, the same ones your apps use.
It chooses the next step, such as which vendors count as late.
Act, check the result, decide again, until the goal is met.
Agents use the app’s Functions as tools. Saved agents run on the EASYProcess scheduler; on-demand agents start when a user asks.
Copilot Studio, n8n or a custom agent connects through the EASYProcess MCP server.
EASYProcess’s built-in service collection.
With the design-time MCP server (see #5).
Build once, use everywhere. Every Function you build is a tool for agents, in-app chat and any AI, and an API for your pages.
“Every weekday at 7:30 AM, find vendors with PO lines unacknowledged for 7 or more days, notify each one, and send me a summary.”
The EASYProcess scheduler starts the agent at 7:30 AM.
Calls the Vendors Function and pages through the results.
Keeps vendors with a PO line open 7 days or more.
Calls NotifySupplier for each vendor, with approval if required.
Sends a summary: who was notified, and any failures.
Easy to change: switch 7 days to 5 by editing one sentence instead of code. Illustrative example.
The model cannot skip it: the approval check lives in EASYProcess, between “AI asked” and “EASYProcess runs it.”
Developers use the built-in EASYProcess tools to add AI prompts directly into business processes. Users get the benefits of AI without worrying about how to use AI.
Uses the built-in EASYProcess tools to add AI prompts to an app.
AI runs inside the everyday business process.
Gets the results without having to learn AI.
Through the design-time MCP server, you build and change the Functions and pages in the project you choose, with Claude or another AI at your side. Your users get a normal, fast EASYProcess app.
Open a project in the EASYProcess IDE and add the pages and Functions AI will work on: existing ones to change, or empty new ones to build.
With AI’s help, you build and change the Functions in the project, using design-time tools like ModifyFunction.
With AI’s help, you build and change the pages in the project with ModifyPage. Each page calls its Functions as REST APIs.
Any AI connects through MCP and works only on what is in the project.
A normal EASYProcess page calling EASYProcess REST APIs. No AI at run time.
Any developer can start building with EASYProcess on day one.
Developers create the project in the EASYProcess IDE and choose which pages and Functions go into it.
For something new, add an empty page or Function to the project. AI builds it out from there.
AI can’t create projects, add objects to a project, or touch anything outside it.
A chatbot inside the IDE, opened from the canvas right where developers build. It works in every environment.
Illustrative example
Secure by design: you control what AI sees, AI runs as the user, JD Edwards stays isolated, production stays protected, and people approve changes. Book a 30-minute demo and see it with your own JDE use case.