Sage 50 and FedEx integration

What moves between them

The primary data flow runs from Sage 50 into FedEx and back. Purchase orders and goods receipts are polled from Sage 50 on a regular schedule, and completed or in-transit shipments trigger tracking lookups in FedEx via the Track API. Shipment tracking events and delivery confirmations are then synced back into Sage 50 as updates to the corresponding goods received notes and purchase order records, so warehouse staff and accounting have a shared view of inbound inventory status. Shipment creation requests generated from Sage 50 purchase orders flow into FedEx using the Ship API, with FedEx tracking numbers written back to Sage 50 for ongoing visibility. Because FedEx has no invoicing REST API, freight invoice reconciliation remains manual via the FedEx Billing Online portal.

How ml-connector handles it

ml-connector runs on a Windows machine with Sage 50 installed and uses the Windows SDK (US .NET SDK or UK SDO COM layer) to authenticate with Windows-local credentials and read purchase orders, goods receipts, and inventory items from the company data files. On the FedEx side, ml-connector stores the OAuth2 API Key and Secret encrypted, caches the bearer token (with a 1-hour TTL to avoid auth rate limits of 3 tokens per second), and refreshes it on 401 responses. When ml-connector creates a shipment from a Sage 50 purchase order, it submits the Ship API request with a transactionId drawn from the PO number for client-side deduplication, since FedEx does not guarantee server-side idempotency. Tracking lookups use the Track API to fetch up to 30 tracking numbers at once, respecting the 1,400-transaction-per-10-second rate limit with exponential backoff. Shipment details (weight, destination, service type) are mapped from Sage 50 PO line items to FedEx shipment fields, and FedEx tracking numbers are written back to Sage 50 purchase order records so the two stay linked. Because Sage 50 has no webhooks, both systems are polled on a fixed schedule tied to your receiving and shipping rhythm. Every record carries an audit trail of what moved, when, and why, and failed shipment creations or tracking updates can be retried.

A real-world example

A mid-sized wholesale distributor runs Sage 50 on a Windows machine for accounting, inventory, and supplier management, and ships packages via FedEx. Before the integration, the receiving staff created purchase orders in Sage 50 when placing vendor orders, then manually tracked inbound shipments in FedEx and later logged delivery dates back into Sage 50 to update goods receipts. Outbound, warehouse staff printed FedEx labels outside of Sage 50 and then had to re-enter tracking numbers into PO records for historical reference. With Sage 50 and FedEx connected, inbound purchase orders trigger FedEx tracking lookups, and delivery events flow back into Sage 50 automatically, so receiving knows package status without switching systems. Outbound shipments are created from sales orders directly to FedEx, and tracking numbers are stored in Sage 50 for traceability. Month-end reconciliation of freight costs is easier because costs and shipments are now linked within the same workflow.

What you can do

• •Poll purchase orders and goods receipts from Sage 50 and create matching FedEx shipments with automatic tracking number linkage.
• •Sync FedEx tracking events and delivery confirmations back into Sage 50 goods received notes so receiving staff have real-time shipment status without leaving the accounting system.
• •Map Sage 50 PO line items to FedEx shipment details (weight, destination, service type) and validate addresses against FedEx standards.
• •Handle Sage 50 Windows SDK authentication (both US .NET and UK SDO editions) and FedEx OAuth2 with token caching to stay within rate limits.
• •Maintain a full audit trail of every shipment created and tracking event synced, with the ability to replay failed updates without duplication.

Questions

How does ml-connector integrate with Sage 50 when it has no cloud REST API?

Sage 50 provides a local Windows SDK (US .NET SDK or UK SDO COM layer) that ml-connector uses to read and write purchase orders, goods receipts, and inventory directly from the company data files. The integration must run on a Windows machine with Sage 50 installed and access to the data files. Authentication is Windows-local username and password plus the application-specific parameters (ApplicationID and CompanyPath for US, DataPath for UK).

What happens if FedEx rate limits are hit during shipment creation or tracking lookups?

ml-connector caches the OAuth2 bearer token with a 1-hour TTL to avoid triggering the 3-tokens-per-second auth limit. For transaction rate limits (1,400 per 10 seconds), ml-connector uses exponential backoff and retries. The Track API can fetch up to 30 tracking numbers in one request, so batching reduces the transaction count. Every rate-limited request is logged with a timestamp so you can see when limits were hit and adjust polling frequency if needed.

Can ml-connector create FedEx shipments from Sage 50 purchase orders automatically?

Yes. ml-connector polls Sage 50 purchase orders on a schedule you set and can trigger FedEx shipment creation for marked orders. It maps Sage 50 PO line items to FedEx shipment fields (weight, destination, service type), uses the PO number as a transactionId for deduplication, and writes the FedEx tracking number back to the Sage 50 record so both systems stay synchronized.