Oracle PeopleSoft and Walmart Marketplace integration

What moves between them

Orders from Walmart Marketplace are polled and matched against open purchase orders in PeopleSoft by item SKU and supplier. When a match is found, ml-connector creates a receipt in PeopleSoft tied to the matched PO, and updates the PeopleSoft inventory record with the order quantity and receipt status. Inventory quantity changes in PeopleSoft are pushed back to Walmart to keep your seller-side stock levels accurate. The integration runs on a polling schedule you define, typically daily or per sales cycle.

How ml-connector handles it

ml-connector stores both OAuth2 credentials encrypted and manages Walmart's 15-minute token refresh automatically, re-requesting a new token when the current one approaches expiry or when a call returns 401. On the PeopleSoft side, it accepts the full customer-specific hostname and port, since there is no Oracle-managed shared base URL, and uses either HTTP Basic Auth or OAuth2 bearer token depending on the PeopleSoft version. Orders are matched to purchase orders using the item SKU and supplier name, and where the match is ambiguous or missing, the record is flagged in the audit trail for manual review. Large data sets are handled with date-range filters on both sides to avoid timeout. Every transaction is logged with full audit trail and can be replayed if a downstream call fails.

A real-world example

A wholesale distributor sells merchandise through Walmart Marketplace while managing procurement and inventory in Oracle PeopleSoft. Before the integration, the warehouse team received Walmart orders separately from PeopleSoft inbound shipments, requiring manual cross-checking to verify that sold items matched open purchase orders and that received quantities updated correctly. When a Walmart order was placed, inventory in PeopleSoft was not automatically decremented, leading to occasional oversells and expedited reshipping. With PeopleSoft and Walmart Marketplace connected, each Walmart order automatically creates a receipt record tied to the source purchase order in PeopleSoft, inventory is updated in real time, and the finance team can reconcile supplier payments against both the PO and the actual sales. The manual matching step is eliminated, and inventory accuracy improves.

What you can do

• •Match Walmart Marketplace orders to Oracle PeopleSoft purchase orders by item and supplier, and create inbound receipts in PeopleSoft.
• •Update PeopleSoft inventory quantities based on Walmart order placements and receipt confirmations.
• •Manage Walmart's 15-minute OAuth2 token refresh and PeopleSoft's per-customer hostname and authentication per version.
• •Handle date-range filtering to process large order and inventory batches without timeout, and surface unmatched orders in the audit trail for manual review.
• •Poll both systems on a schedule tied to your supply cycle, with retries and full transaction logging on every order and inventory change.

Questions

How does the integration handle PeopleSoft's lack of webhooks?

ml-connector polls both PeopleSoft and Walmart Marketplace on a schedule you define, checking for new Walmart orders and updated inventory at regular intervals. This approach works well for daily or near-real-time syncs and avoids the need for PeopleSoft to expose a webhook endpoint. Each poll captures orders and inventory changes since the last run using date-range filters.

What happens when a Walmart order does not match any open purchase order in PeopleSoft?

Unmatched orders are flagged in the audit trail with the reason why the match failed, such as missing supplier name or no open PO for that item. ml-connector does not create a blind receipt without a PO match. The warehouse team can review the flagged record and either create the PO manually in PeopleSoft or investigate whether the order belongs to a different supplier.

How are Walmart's 15-minute token expiries handled?

ml-connector requests a new OAuth2 token 2 minutes before expiry and caches it. If a call returns 401 Unauthorized, ml-connector immediately refreshes the token and retries the request. This ensures uninterrupted syncing and no failed orders due to expired credentials.