IFS Cloud and ServiceNow integration

What moves between them

Supplier invoices and purchase orders flow from IFS Cloud into ServiceNow. ml-connector polls IFS OData on a schedule, retrieves new and modified supplier invoices (PostingProposalHeadSet) and purchase orders (PurchaseOrderSet), enriches them with vendor and GL dimension data, and stages them in ServiceNow's sn_apo_invoice and proc_po tables for AP and procurement workflows. GL accounts and cost centers are pre-mapped before staging so finance coding is valid on arrival. The flow is one-way from IFS to ServiceNow; financial postings are not written back to IFS.

How ml-connector handles it

ml-connector stores both OAuth credential sets encrypted and refreshes tokens before expiry on each call. For IFS Cloud, it constructs the tenant-specific OData URL and passes the OAuth 2.0 bearer token on every request, respecting the 5000-element page limit and 1000-request-per-minute rate ceiling with exponential backoff on HTTP 429 responses. For ServiceNow, it validates that the target instance has the Source-to-Pay module enabled and the service account holds the required procurement and AP roles. Supplier invoices are enriched with vendor names and cost centers before staging; purchase orders carry the matching GL account and cost center from IFS so they land on valid ServiceNow dimensions. Because IFS enforces ETag-based optimistic concurrency on updates, ml-connector reads each record once to capture its ETag before any mutation. Every record is deduplicated before write using a combination of invoice or PO number and creation date, preventing duplicate staging if a job re-runs. Retries use exponential backoff with jitter, and a full audit trail tracks every read and stage operation.

A real-world example

A medium-scale manufacturing company runs IFS Cloud for global production and procurement and ServiceNow for IT and compliance. The procurement team receives supplier invoices in IFS, but AP processing and three-way matching happen in ServiceNow. Before the integration, invoices were exported from IFS, copied into a spreadsheet, and re-entered into ServiceNow's staging tables by hand, causing mismatches between PO commitments and AP balances and delaying month-end close. With IFS and ServiceNow connected, invoices land in ServiceNow staging automatically on receipt, matched to the correct cost center and GL account from IFS, ready for the three-way match. The procurement team works from ServiceNow, AP reconciliation happens in IFS, and both systems stay in agreement.

What you can do

• •Poll IFS Cloud for supplier invoices and purchase orders on a schedule, with automatic OAuth token refresh.
• •Stage supplier invoices and POs in ServiceNow Accounts Payable and Procurement tables with GL account and cost center pre-mapped from IFS.
• •Validate GL accounts and cost centers against the IFS instance before staging to prevent downstream errors.
• •Respect IFS rate limits and page size constraints, with exponential backoff on throttle responses.
• •Deduplicate invoices and POs across job runs using invoice number, PO number, and creation date to prevent staging errors.

Questions

Which direction does data move between IFS Cloud and ServiceNow?

The main flow is from IFS Cloud into ServiceNow. Supplier invoices and purchase orders are polled from IFS OData and staged in ServiceNow's Accounts Payable and Procurement tables for approval and posting. Financial postings and payments are not written back to IFS.

How does ml-connector handle IFS Cloud's OAuth 2.0 authentication and token expiry?

ml-connector stores both IFS and ServiceNow OAuth credentials encrypted. It refreshes the IFS token proactively before expiry on each poll cycle and the ServiceNow token before staging operations, so token expiry does not interrupt the sync. If a call returns a 401, the token is refreshed and the call is retried.

What happens if an invoice or PO contains a GL account or cost center that does not exist in ServiceNow?

ml-connector validates GL accounts and cost centers against the IFS configuration before staging. If a dimension is missing, the record is flagged in the audit trail with the validation error and not staged, allowing the finance team to correct the master data in IFS before retry.