Oracle JD Edwards and Cleo integration

What moves between them

Purchase orders and invoices flow from Oracle JD Edwards EnterpriseOne into Cleo Integration Cloud. ml-connector polls JD Edwards at a configurable interval for new and updated purchase orders (F4301/F4311), accounts payable documents (F0411), and supporting reference data such as suppliers (F0401) and GL accounts (F0901). Once retrieved, those records are transformed into EDI format (typically 850 for purchase orders, 810 for invoices) or JSON payloads and delivered to Cleo via file upload or API action. Supplier master records and GL dimensions flow both directions so that trading partner mappings and cost allocations remain aligned. Cleo holds the data until a tenant administrator routes it onward to specific suppliers or EDI networks.

How ml-connector handles it

ml-connector stores JD Edwards AIS Server credentials (host, port, username, password) encrypted and maintains a persistent login session, reusing the same token across multiple requests until it expires, at which point it re-authenticates automatically. Since JD Edwards session tokens have a 30 to 60 minute lifetime and authentication requires a valid user license, ml-connector tracks token expiry proactively and refreshes before timeout; an HTTP 444 response signals forced re-authentication. For Cleo, ml-connector uses OAuth 2.0 password grant flow to obtain a Bearer token and refreshes it when a call returns 401. ml-connector polls JD Edwards on a cadence you define (typically daily or weekly for procurement), tracking the last-polled timestamp on UPMJ (updated timestamp for AP) or DGJ (GL date) to avoid re-fetching identical records; results are paginated with maxPageSize=100. Purchase order and invoice details are mapped to Cleo's folder structure and uploaded as EDI files or API-compatible JSON payloads, with supplier master records mapped between systems so that cost centers and GL accounts in JD Edwards correspond to Cleo trading partner identities. Because JD Edwards is on-premises and may have an IP allowlist, the connector's egress IP addresses must be whitelisted by your JD Edwards administrator. ml-connector does not write transactional data back to JD Edwards; all writes from Cleo go through configured CIC Studio actions and do not flow via the REST API.

A real-world example

A mid-sized manufacturing company runs Oracle JD Edwards EnterpriseOne for production, procurement, and finance across three plants and a central purchasing team. They use Cleo Integration Cloud to connect their suppliers and route EDI transactions. Before integration, the purchasing team created purchase orders in JD Edwards, exported them as text files, manually re-keyed them into Cleo's portal, and tracked supplier acknowledgments outside both systems. With Oracle JD Edwards and Cleo connected, each purchase order and change order created in JD Edwards automatically flows to Cleo as an 850 EDI transaction, supplier acknowledgments are logged against the original JD Edwards PO, and the purchasing team has a single pane of visibility across creation, transmission, and partner response. Manual re-keying is eliminated, and month-end reconciliation is faster because supplier master records stay synchronized.

What you can do

• •Poll JD Edwards purchase orders, invoices, and AP transactions on a schedule and deliver them to Cleo as EDI files or API payloads.
• •Map JD Edwards supplier master and GL accounts to Cleo trading partner and cost center identities, keeping reference data aligned in both directions.
• •Manage JD Edwards session token lifecycle automatically, refreshing before expiry and handling the 30 to 60 minute token lifetime transparently.
• •Authenticate with OAuth 2.0 on the Cleo side and JD Edwards session tokens on the source side, with encrypted credential storage and proactive token refresh.
• •Maintain a full audit trail on every purchase order, invoice, and transaction moved between systems, with the ability to replay records if downstream delivery fails.

Questions

Does data flow both directions between Oracle JD Edwards EnterpriseOne and Cleo?

The primary flow is JD Edwards into Cleo. Purchase orders, invoices, and AP documents flow from JD Edwards to Cleo for EDI translation and supplier routing. Supplier master records and cost centers are aligned in both directions so trading partner mappings remain synchronized. Cleo does not write financial data back into JD Edwards; all reverse flows use Cleo's configured CIC Studio actions.

How does ml-connector handle JD Edwards' on-premises infrastructure and lack of webhooks?

JD Edwards has no native outbound webhooks and is on-premises, so ml-connector polls the AIS Server on a schedule you define (typically daily or weekly), tracking the last-updated timestamp on AP and GL tables to avoid re-fetching. The AIS Server URL and authentication credentials are stored encrypted per customer, and ml-connector manages the session token lifecycle automatically, refreshing before the 30 to 60 minute expiry window. If your JD Edwards infrastructure has an IP allowlist, the connector's egress addresses must be whitelisted.

What authentication does ml-connector use for each system?

For Oracle JD Edwards EnterpriseOne, ml-connector uses the AIS Server session token, obtained by posting username and password to the tokenrequest endpoint. For Cleo Integration Cloud, it uses OAuth 2.0 password grant flow to exchange credentials for a Bearer token. Both credential sets are stored encrypted, and tokens are refreshed automatically: JD Edwards tokens refresh before the 30 to 60 minute expiry, and Cleo tokens refresh when a 401 response is received.