Oracle PeopleSoft and SAP Ariba integration

What moves between them

The main flow runs from SAP Ariba into Oracle PeopleSoft. ml-connector reads Ariba purchase order and invoice header data through reporting jobs and posts the matching purchase orders and AP vouchers into PeopleSoft through SOAP Component Interfaces, so the procurement documents land in the ERP for payment without re-keying. Supplier data is aligned in both directions: the PeopleSoft vendor master feeds Ariba supplier records through the Supplier Data API, and registration and qualification status is read back from Ariba. Reporting jobs are filtered by an updated-date window, and both reads and the supplier sync run on a polling schedule, typically every several minutes to hourly, with a stored high-water mark so each cycle only picks up new and changed records.

How ml-connector handles it

ml-connector stores both credential sets encrypted. For Ariba it fetches an OAuth2 client-credentials token, caches it and refreshes before the one-hour expiry, sends the static apiKey header on every call, and appends the realm query parameter to each request. For PeopleSoft it accepts the customer-specific base URL and Integration Broker node name, sends Basic Auth with the OPRID and password, and calls the SOAP Component Interfaces for writes since the delivered REST endpoints are read-only inquiry. Ariba bulk reads use the submit-poll-download job pattern, so the connector submits a reporting job, polls until it completes, and pages through results with the pageToken cursor. Because neither system pushes events for this use case, both sides are polled on a schedule rather than waiting for a webhook. Suppliers are mapped first so every voucher and purchase order references a vendor that already exists in PeopleSoft, and PeopleSoft object keys such as voucher and PO numbers act as natural idempotency identifiers so a retry does not create a duplicate. Edge cases are handled directly: Ariba reporting jobs cap a filter window at one year so initial backfill is chunked into annual jobs, Ariba job submission is rate limited so the connector watches the remaining-calls header and backs off before a 429, and PeopleSoft returns HTTP 401 on bad credentials and an inactive service operation returns a generic error, so the connector surfaces both clearly. Every record carries a full audit trail and can be replayed if a downstream call fails.

A real-world example

A mid-sized manufacturer with around 1,500 employees runs Oracle PeopleSoft Financials for accounts payable and the general ledger and uses SAP Ariba for procurement and supplier invoicing across several sites. Before the integration, the AP team exported invoice and purchase order reports from Ariba and keyed the vouchers into PeopleSoft by hand, which slowed payment, introduced typos, and left supplier records out of step between the two systems. With Oracle PeopleSoft and SAP Ariba connected, Ariba invoice headers and purchase orders flow into PeopleSoft as vouchers and POs automatically, and the vendor master stays aligned in both directions. The AP team stops re-keying, invoices reach payment faster, and supplier data matches across procurement and finance.

What you can do

• •Post SAP Ariba invoice headers into Oracle PeopleSoft as AP vouchers through the AP_VOUCHER Component Interface.
• •Move SAP Ariba purchase orders into Oracle PeopleSoft purchase orders through the PO_CI Component Interface.
• •Keep SAP Ariba supplier records aligned with the PeopleSoft vendor master and read registration and qualification status back.
• •Bridge Ariba OAuth2 with its apiKey and realm against PeopleSoft Basic Auth and Integration Broker node name.
• •Poll both systems on a schedule with a high-water mark, idempotent writes keyed on PeopleSoft object numbers, retries, and a full audit trail.

Questions

Which direction does data move between Oracle PeopleSoft and SAP Ariba?

The main flow is SAP Ariba into Oracle PeopleSoft. Purchase orders and invoice headers move from Ariba into PeopleSoft as POs and AP vouchers, while supplier data is aligned in both directions. The PeopleSoft vendor master feeds Ariba supplier records, and registration and qualification status is read back from Ariba.

How does the integration write into PeopleSoft when its REST APIs are read-only?

PeopleSoft's delivered REST inquiry endpoints only read invoice, payment, and requisition status. For writes, ml-connector uses the SOAP Component Interfaces such as AP_VOUCHER and PO_CI exposed by Integration Broker. The PeopleSoft voucher and PO numbers act as natural keys, so re-sending a record that already exists is rejected rather than duplicated.

Does this integration use webhooks or polling?

It uses polling, because neither system offers a simple outbound webhook for this use case. Ariba data is pulled through its submit-poll-download reporting jobs filtered by an updated-date window, and PeopleSoft is read on a schedule unless an admin configures Integration Broker async routing. ml-connector keeps a high-water mark so each cycle only picks up new and changed records.