Oracle E-Business Suite and Tradeshift integration

What moves between them

AP invoices and purchase orders flow from Oracle E-Business Suite into Tradeshift as UBL XML documents, published to the supplier network for procurement and payment workflows. Receipts, remittance advice, and other inbound documents flow from Tradeshift back into EBS, where they are written to open interface tables for import into the financial and procurement base tables. The integration polls EBS on a daily or weekly schedule and polls Tradeshift for inbound documents using the changedAfter timestamp filter.

How ml-connector handles it

ml-connector stores EBS ISG hostname, port, Basic auth credentials or session token in encrypted storage, and renews the session token when calls return 401 or when the token nears expiry. It also stores Tradeshift OAuth 1.0a credentials (consumer key, consumer secret, access token, token secret) encrypted and includes the X-Tradeshift-TenantId header on every call. When writing to EBS, ml-connector posts AP invoices and POs to the appropriate open interface tables via REST, then triggers the associated concurrent programs to import those records into the base tables, which can take minutes to hours depending on validation rules. When reading from Tradeshift, it polls the documents endpoint with a changedAfter filter and transforms UBL XML back into EBS GL posting dimensions and AP line details, mapping Tradeshift document types to the correct EBS interfaces. Every record carries a job ID for idempotency so concurrent retries do not create duplicates, and the full audit trail preserves the original document identifiers from both systems.

A real-world example

A mid-market manufacturing company runs Oracle E-Business Suite for procurement, AP, and GL. Their procurement team creates purchase orders in EBS and has been manually uploading them to Tradeshift to notify suppliers, then downloading receipts and remittance advice from Tradeshift back into a spreadsheet before posting them into EBS AP. With EBS and Tradeshift connected, new purchase orders post to the Tradeshift network automatically within hours of approval in EBS, suppliers can see them without manual uploads, and remittance advice and receipt confirmations flow back into EBS GL and AP nightly. The accounting team no longer manually reconciles AP and PO data across two systems, and suppliers get faster visibility into order status.

What you can do

• •Publish Oracle E-Business Suite purchase orders and AP invoices to Tradeshift as UBL XML documents on a schedule.
• •Receive Tradeshift receipts, remittance advice, and credit notes and post them into EBS AP and GL open interface tables.
• •Authenticate EBS via HTTP Basic and session tokens, and Tradeshift via OAuth 1.0a with tenant ID headers.
• •Map EBS GL code combinations, vendors, and invoice lines to UBL document structure and Tradeshift network fields.
• •Poll both systems on a schedule with automatic session renewal, job ID deduplication, and a full audit trail on every document.

Questions

What documents move between Oracle E-Business Suite and Tradeshift?

Purchase orders and AP invoices move from EBS into Tradeshift so suppliers can see them on the B2B network. Receipts, remittance advice, credit notes, and despatch advice flow from Tradeshift back into EBS, where ml-connector posts them to open interface tables for import into AP and procurement. Other documents such as quotations and network connection profiles can also be exchanged as needed.

How does ml-connector handle the different authentication schemes?

EBS uses HTTP Basic authentication (username and password) or a session token obtained via the login endpoint, with application context headers for responsibility and organization. Tradeshift uses OAuth 1.0a with a consumer key, consumer secret, and access token, plus the X-Tradeshift-TenantId header. ml-connector stores both credential sets encrypted and applies the correct auth method to each system on every call, renewing EBS session tokens when they expire.

Does Tradeshift support real-time document push, or does ml-connector poll?

Tradeshift offers optional continuous event streams via protobuf for registered plugins, but the standard polling approach uses the changedAfter timestamp filter on the documents endpoint. ml-connector polls both systems on a daily or weekly schedule so that new documents from either side are synced within your chosen cadence, and every document is tracked by job ID to prevent duplicate posts if a retry occurs.