● JD Edwards EnterpriseOne · Applied AI · Enterprise Architecture

Making enterprise ERP intelligent, without making it fragile.

I'm Mohd Shahid Equbal, Manager of ERP Development at Connection, a Fortune 1000 technology solutions company. For 20+ years I've built and run JD Edwards systems. Today I research how AI should be wired into them, and I coined the term AI Integration Debt to name what goes wrong when it isn't.

Read the research → Talks & community

20+

years in JD Edwards EnterpriseOne, from developer to architecture lead

new term coined: AI Integration Debt, published July 2026

Quest Oracle Community sessions across InFocus and Blueprint 4D

academic databases in an ongoing PRISMA-ScR scoping review

F0411 · the ledger nobody is keeping

AI Integration Debt

Every ERP team is under pressure to ship AI. The integrations go in fast, and the cost of maintaining, auditing and updating them builds up quietly. My research gives that cost a name, a structure, and a way to measure it.

Definition

The accumulated cost of expedient, undocumented, or ungoverned AI service integrations into enterprise ERP systems, showing up as increased maintenance burden, reduced reliability, and governance risk over time.

First published July 3, 2026 · blog.meshahid.ai

Long system lifespans

ERP platforms run for decades. Each integration must outlive several generations of AI models and APIs.

Deep process coupling

A faulty AI call in procure-to-pay doesn't produce a bad prediction. It produces a bad purchase order.

Layers of customization

AI sits on top of years of bespoke business logic, so debt compounds with every layer.

Governance gaps

Change management covers ERP code but rarely model versions, drift, or rollback. That gap is the debt.

Conference paper · in preparation

A Conceptual Taxonomy and Provisional Assessment Profile

A five-dimension taxonomy of AI Integration Debt in ERP systems, with the AIDI assessment instrument. Built to profile risk rather than collapse it into a single score.

Scoping review · ongoing

PRISMA-ScR review across five databases

Scopus, Web of Science, IEEE Xplore, ACM Digital Library and AIS eLibrary, searched across technical debt in ML, AI in ERP, legacy ERP debt, and AI governance.

Essay · published

Your ERP Has a Hidden AI Problem

The essay that introduced the term, with field patterns like the orphaned integration, the undocumented dependency, and the governance void. Read it →

Framework · InFocus 2026

The 5-Layer AI-Augmented Development Model

A practical model for bringing AI into the JD Edwards development lifecycle under control. It keeps deterministic automation clearly separate from real AI assistance, and puts governance in the design from the start.

Intent

What business outcome are we changing, and in which E1 objects?

Every change starts with a clear statement of the business requirement and the JD Edwards objects it touches. AI helps sharpen the requirement and map it to applications, UBEs and tables, but a person owns the intent.

AI-assisted E1 logic design from plain-language requirements
Early impact mapping across applications, reports and tables
Human sign-off before anything is generated

From the session “The AI-Augmented JD Edwards Developer: A Practical Framework for the Next Era of ERP Development” · InFocus 2026, Denver.

Experience · two decades on one platform

Builder, consultant, architect, leader

My career has covered every side of the JD Edwards table: writing the code, advising clients as a consultant, and owning architecture and delivery inside the business.

Current

Manager, ERP Development

Connection (NASDAQ: CNXN) · Merrimack, NH

Lead JD Edwards EnterpriseOne development for a Fortune 1000 technology solutions provider.
Built a Python release automation pipeline that ties JDE releases to Jira workflow gates, database checks and stakeholder notifications.
Designed an AI-ready invoice platform and a modern rebate architecture, both presented at Blueprint 4D 2026.
Received the Achievers Club Award, with a rare dual nomination from both Sales and Procurement.

Earlier

Enterprise architecture leadership

Customer-side ERP organizations

Owned ERP architecture decisions from inside the business, where integrations have to keep working for years.
Learned firsthand how integrations become load-bearing, the insight behind AI Integration Debt.

Earlier

Consultant

IBM

Delivered JD Edwards solutions for enterprise clients.

Foundation

JD Edwards Developer

Where it started

Built applications, UBEs and integrations in EnterpriseOne. The hands-on base that everything since has been built on.

Speaking · Quest Oracle Community

On stage with the JD Edwards community

I speak where JD Edwards practitioners gather, on the practical side of bringing AI into ERP development and operations. At Blueprint 4D 2026 I gave two sessions, including a co-presentation with Jason Gendreau, Director of ERP Applications Development at Connection.

InFocus 2026DenverSession P-053116

The AI-Augmented JD Edwards Developer: A Practical Framework for the Next Era of ERP Development

A 60-minute Insights session walking through real scenarios, from AI-assisted E1 logic design to impact analysis, test design, runbooks, and a governed release pipeline.

Intent L1Generation L2Validation L3Execution L4Governance L5

Blueprint 4D 2026

Building an AI-ready invoice platform

How to structure invoice processing in JD Edwards so AI can be added safely and governed from day one.

Blueprint 4D 2026

Modern rebate architecture

Rethinking rebate processing on JD Edwards with a cleaner, more maintainable architecture.

InFocus 2025

Speaker, JD Edwards InFocus 2025

Presented at InFocus, the largest annual gathering of the JD Edwards user community.

Essay · forthcoming

What InFocus 2026 tells us about the next era of JDE

A practitioner's reflections on the conference and why sharing what we learn matters for the whole community.

Quest Oracle Community

JD Edwards Advisory Board

Member of the advisory board serving the JD Edwards user community.

Oracle

Oracle ACE Associate

Recognized in 2026 for JD Edwards EnterpriseOne expertise.

IEEE New Hampshire Section

Webmaster & Executive Committee

Runs the section's web presence, including the K-12 grant program.

Research community

Peer reviewer

Reviewing for IEEE conference tracks through CMT.

Credentials · verified, current

Engineering grade, AI certified, academically grounded

Professional recognition from engineering bodies, hands-on certifications in AI and data platforms, and an education that runs from manufacturing engineering to machine learning.

Professional recognition

IEEE Senior Member2026
Member, IET (MIET)2026
Oracle ACE Associate2026
CIPS I.S.P. / ITCPCanada
SFIA Level 5confirmed

AI, cloud & data

Azure AI Engineer AssociateAI-102
Azure AI FundamentalsAI-900
Databricks Data Engineer AssociateDatabricks
AWS Cloud PractitionerAWS

Education

MS Computer Science (ML)Georgia Tech · in progress
Executive MBAUniv. of New Haven
B.Tech, Manufacturing Science & EngineeringIIT Kharagpur

Writing · blog.meshahid.ai

Notes from the intersection of ERP and AI

Essays on AI Integration Debt, governed AI in JD Edwards, and what practitioners are learning in the field. The series continues with a measurement framework and governance model.

Read the blog →

Now · autumn 2026

What I'm working on

AI Integration Debt taxonomy paper · preparing for submission
Scoping review of AI, ERP and technical debt · screening and adjudication
Georgia Tech OMSCS · CS 7641 Machine Learning, Fall 2026
JD Edwards Advisory Board · community programming

F0101 · address book

Let's talk about AI in your ERP.

I'm glad to hear from conference organizers, editors, researchers and ERP leaders working on the same problems. Speaking invitations, research collaboration and advisory conversations are all welcome.

shahid@meshahid.ai

Blog

blog.meshahid.ai ↗

ORCID

0009-0002-2083-3600 ↗