SlackJD Edwards EnterpriseOne Agent view

How to add Slack to JD Edwards EnterpriseOne

Short answer

Create a Slack app that calls JD Edwards through the AIS Server and Orchestrator. Outbound alerts come from scheduled orchestrations or Real Time Events, slash commands run read-only lookups, and interactive buttons trigger approved actions. An on-prem JDE is reached through an outbound relay rather than by opening the network.

Get this built on your JD Edwards EnterpriseOne

Slack tends to appear in JD Edwards shops where engineering, supply chain or a newer business unit works alongside an older ERP footprint. The people who use Slack all day are rarely the ones fluent in JDE application names like P4210 or P4310, so questions bounce between channels and a few experts.

A well scoped Slack app turns those experts' repeat answers into commands. Instead of asking in a channel whether a sales order has shipped, a user types a slash command and the app returns the status pulled live from EnterpriseOne.

The pattern is simple but the guardrails matter. Slack identities are not JDE identities, public channels leak data, and any write action needs the same controls it would have in the JDE web client. This page covers how to wire it using the pieces JDE already has.

What teams use it for

Slash command order and PO lookup

/jde order 40123 returns status, ship date and held reasons from sales order tables; /jde po 5521 returns open lines and promised dates.

Shortage alerts to a supply channel

Post critical shortages where open demand exceeds supply for an item, with item number, branch/plant and next planned receipt.

Interactive approval messages

Send requisition or change-order approvals with Approve and Reject buttons to the approver's direct message.

Engineering change notices

Announce when a bill of material (F3002) or routing is changed, so engineering and production both see it.

Production incident threads

When a work order is flagged as late or on hold, open a thread in the line channel with a link to the JDE form.

Customer escalation lookups

Let account managers check open orders and backorders for a customer from Slack without a JDE license seat.

JD Edwards EnterpriseOne integration surface

AIS Server REST endpoints, especially Orchestrator calls published for each named action; the AIS base URL, security and token behavior depend on your Tools release.
Orchestrator Studio schedules and notifications to generate outbound events without custom polling code.
Real Time Events where available, delivering change events to a listener that your relay can transform into Slack posts.
Reading data through AIS data requests on F4211, F4311, F3111, F41021 (item location) and F0101; direct SQL reads against the production database are not recommended.
JDE security records for application and row level access, which should shape what each Slack user can see.
UBE reports and One View Reporting exports for batch style summaries that you push to a channel on a schedule.

How it works

1
Create the Slack app
Define slash commands, bot scopes and interactivity URLs. Use Socket Mode if you do not want an inbound public endpoint for the app, which suits on-prem sites.
2
Build the orchestrations
Create small read orchestrations first (order, PO, item availability) and add write ones only after review. Return compact JSON shaped for display.
3
Run a relay in your network
The relay holds the Slack tokens, calls AIS over the internal network, and formats Block Kit messages. Slack never talks to JDE directly.
4
Resolve identity
Map Slack user IDs to JDE users in an owned table. For read commands you may allow a role-based default; for writes require the exact mapping.
5
Handle interactions safely
Verify Slack signatures, include a correlation ID in each button payload, and make the action idempotent so double clicks are harmless.
6
Control channels
Post sensitive details only to private channels or DMs. Keep the default response ephemeral so only the requester sees pricing and customer data.

Example flows

Slash command for order status

Trigger: A user types a slash command with an order number.

The relay verifies the Slack signature and finds the mapped JDE user.
It calls the GetOrderStatus orchestration through AIS.
The response is formatted as an ephemeral message with status, ship date and a link to the JDE form.

Result: A salesperson answers a customer in under a minute without hunting for a JDE seat.

Shortage alert with acknowledge

Trigger: A scheduled orchestration identifies items whose projected availability goes negative within the planning fence.

The relay groups items by planner and posts to the supply channel.
The planner taps Acknowledge or Assign to me.
The relay stores the acknowledgement and optionally writes a note back through a controlled orchestration.

Result: Shortages have an owner and a visible response time instead of living in email threads.

Change order approval

Trigger: An engineering change order awaits approval in JDE workflow.

The relay sends a message with item, reason and affected work orders to the approver.
The approver reviews the linked JDE form and taps Approve.
The approval orchestration runs under the mapped identity.

Result: Faster sign-off with the decision recorded in JDE.

Keeping data on-prem

Slack is a cloud service, but JDE can remain entirely behind your firewall. The cleanest design is an on-prem relay using Slack Socket Mode, so it connects outward and no inbound port is opened. The relay calls AIS internally and sends only the text you choose to Slack. If you must keep certain data out of third-party systems, restrict messages to identifiers, statuses and links, and keep detail in JDE. Natural-language commands can be interpreted by a locally hosted open-weight model that only selects among your fixed orchestrations.

Security checklist

Verify every Slack request signature and rotate bot and signing secrets on a schedule.
Use ephemeral replies and private channels for any commercially sensitive data.
Grant the integration user JDE security for only the orchestrations it needs.
Log who ran which command and which orchestration was called, with timestamps.
Review Slack workspace retention and export settings, since messages become a second record of ERP data.

Ways to get it done

No-code automation platform with a JDE REST step

Effort: Days.

Fine for simple alert posts when AIS is already reachable.

Custom Slack app and relay in-house

Effort: A few weeks for a first set of commands.

Best for slash commands, mapped identity and private deployment.

Partner build with Netray

Effort: Pilot per use case, then extension.

Good when you also want natural-language questions answered over JDE data by a private model.

Pitfalls

!Returning pricing or customer-specific terms in a public channel.
!Allowing free-text commands to map to arbitrary AIS calls rather than a fixed list.
!Rate-limiting problems when posting large alert batches one message at a time.
!No handling for AIS being down during a JDE package deployment or restart window.
!Treating Slack button clicks as authentication without verifying the mapped user.

Frequently asked questions

Can Slack talk to JD Edwards directly?

Not on its own. Slack calls your app or relay, which then calls the AIS Server and Orchestrator endpoints. The relay is where you enforce identity mapping, formatting and access rules.

Do I need to open an inbound port to my JDE network?

No. Use Slack Socket Mode or an outbound-initiated tunnel from the relay, and keep AIS accessible only on your internal network. This is the usual approach for on-prem installations.

Which JDE actions are safe to expose in Slack?

Start with read-only lookups and acknowledgements. Add approvals for low-risk items next. Keep financial postings, pricing changes and master data updates in JDE unless there is a reviewed, audited flow.

Will this work with JD Edwards World?

World lacks AIS and Orchestrator, so you would typically use database exports, flat files or another adapter. That is a different and more limited design than EnterpriseOne.

How do I stop sensitive data ending up in Slack history?

Send ephemeral responses, restrict channels, and limit message content to what the audience should see. For the most sensitive items send a link back to JDE instead of the data itself.

More for JD Edwards EnterpriseOne

AI agents ChatGPT Claude MCP server AI chatbot Power BI WhatsApp Microsoft Teams Email automation Document OCR Mobile app

AI for ERP

AI for JD Edwards EnterpriseOne, Built on Orchestrator and AIS

AI for ERP

AI for JD Edwards Inventory Management and Shop Floor Manufacturing

Want Slack working on your JD Edwards EnterpriseOne?

A free 30 minute assessment: we map the integration path for your version and setup, and where it should run.

Get a free assessment

All integrations

Try it here, free

JD Edwards Orchestrator and AIS REST Call Generator

Open full tool

Fill in your AIS host, orchestration name and inputs to get a working Python requests script and curl commands that call a JD Edwards EnterpriseOne 9.2 orchestration.

Fill in

AIS Server base URL

Protocol, host and port of the AIS Server, without /jderest.

Orchestration name

The name exactly as saved in Orchestrator Studio.

Orchestrator API version

Authentication

False = HTTP Basic auth on the orchestrator call. True = request an AIS token first, send it in the jde-AIS-Auth header, then log out.

EnterpriseOne user

Environment (token mode)

Role (token mode)

Device name

Orchestration inputs (JSON)

Use the input names defined on the orchestration.

Generated code

Free, optional. Reopen it on any device.

Download:

Python (requests)

# Call JD Edwards orchestration JDE_ORCH_AddSalesOrder through AIS (v3)
# pip install requests
import json
import os

import requests

AIS = 'https://ais.example.com:9300/jderest'
ORCH = 'JDE_ORCH_AddSalesOrder'
USER = 'INTEGUSER'
PASSWORD = os.environ.get('JDE_PASSWORD', '')  # never hard-code the password
ENV = 'JDV920'
ROLE = '*ALL'
DEVICE = 'integration-host-01'
USE_TOKEN = False
VERIFY = True  # or the path to your internal CA bundle, e.g. '/etc/ssl/certs/corp-ca.pem'

inputs = json.loads(r'''{"Customer": "4242", "Branch": "30"}''')

session = requests.Session()
session.headers.update({'Content-Type': 'application/json', 'Accept': 'application/json'})
token = None
auth = None

if USE_TOKEN:
    # (1) POST /jderest/v2/tokenrequest, token is returned in userInfo.token
    login = session.post(
        AIS + '/v2/tokenrequest',
        json={'username': USER, 'password': PASSWORD, 'environment': ENV,
              'role': ROLE, 'deviceName': DEVICE},
        timeout=60, verify=VERIFY)
    login.raise_for_status()
    token = login.json()['userInfo']['token']
    session.headers.update({'jde-AIS-Auth': token, 'jde-AIS-Auth-Device': DEVICE})
else:
    # Basic auth: AIS uses its configured default environment and role
    auth = (USER, PASSWORD)

try:
    # (2) POST /jderest/v3/orchestrator/{name} with the orchestration inputs
    resp = session.post(AIS + '/v3/orchestrator/' + ORCH,
                        json=inputs, auth=auth, timeout=120, verify=VERIFY)
    resp.raise_for_status()
    print(json.dumps(resp.json(), indent=2))
finally:
    if token:
        # (3) POST /jderest/v2/tokenrequest/logout ends the AIS session
        session.post(AIS + '/v2/tokenrequest/logout', json={'token': token},
                     timeout=30, verify=VERIFY)

Same call with curl (Basic auth)

curl -s -X POST 'https://ais.example.com:9300/jderest/v3/orchestrator/JDE_ORCH_AddSalesOrder' \
  -u 'INTEGUSER:YOUR_PASSWORD' \
  -H 'Content-Type: application/json' \
  -H 'Accept: application/json' \
  -d '{"Customer": "4242", "Branch": "30"}'

curl with an AIS token

TOKEN=$(curl -s -X POST 'https://ais.example.com:9300/jderest/v2/tokenrequest' \
  -H 'Content-Type: application/json' \
  -d '{"username":"INTEGUSER","password":"YOUR_PASSWORD","environment":"JDV920","role":"*ALL","deviceName":"integration-host-01"}' \
  | jq -r '.userInfo.token')

curl -s -X POST 'https://ais.example.com:9300/jderest/v3/orchestrator/JDE_ORCH_AddSalesOrder' \
  -H 'Content-Type: application/json' \
  -H "jde-AIS-Auth: $TOKEN" \
  -H 'jde-AIS-Auth-Device: integration-host-01' \
  -d '{"Customer": "4242", "Branch": "30"}'

curl -s -X POST 'https://ais.example.com:9300/jderest/v2/tokenrequest/logout' \
  -H 'Content-Type: application/json' \
  -d "{\"token\":\"$TOKEN\"}"

Paths follow the Oracle JD Edwards EnterpriseOne 9.2 AIS REST API reference: POST /jderest/v2/tokenrequest (token in userInfo.token), POST /jderest/v3/orchestrator/{orchestration} (token header jde-AIS-Auth), POST /jderest/v2/tokenrequest/logout with a token body field. Check that your Tools release supports v3 and Basic auth on the orchestrator endpoint, and that the user is authorised to the orchestration in Security Workbench.

Get your AIS integration reviewed

Send us the orchestration and your target client (Python, Power Automate, MES) and we will check auth, error handling and session cleanup against your Tools release.

Optional. The tool and every export work without this. No spam.

ArchRay builder

Design the on-prem AI stack for JD Edwards EnterpriseOne: model, GPUs, RAG pipeline, guardrails

See memory, speed and cost as you swap parts, then export the diagram. Free, in your browser.

Open the builder

New to JD Edwards EnterpriseOne? Start with the JD Edwards EnterpriseOne guide and toolkit, or plan with Training Path Builder, Upgrade Path Planner, ERP Table and Field Explorer.